Content uploaded by Alexander Curth
Author content
All content in this area was uploaded by Alexander Curth on Mar 09, 2024
Content may be subject to copyright.
Construction and Building Materials 421 (2024) 135714
Available online 8 March 2024
0950-0618/© 2024 Elsevier Ltd. All rights reserved.
3D printing earth: Local, circular material processing, fabrication methods,
and Life Cycle Assessment
Alexander Curth
*
, Natalie Pearl , Angelica Castro-Salazar , Caitlin Mueller, Lawrence Sass
Massachusetts Institute of Technology, Architecture – Computation, Cambridge, USA.
ARTICLE INFO
Keywords:
3D printed earth
3D printed earth formwork
Additive manufacturing
Earth architecture
Soil engineering
ABSTRACT
Additive manufacturing with earth is an emerging, though largely uncharacterized, approach to fabricating low
embodied carbon structures. It is critical to establish methods for processing 3D printed, locally sourced earthen
materials in different environments to validate large-scale earthen additive manufacturing as a tool to address a
growing global need for housing and climate-resilient architecture. We present a set of reproducible design
guidelines for sourcing, processing, and characterizing locally sourced earthen materials. Soil type, moisture and
ber content, particle size distribution, and unconned compressive strength are determined. Additionally, novel
bridging, cantilevering, and hydrostatic pressure (formwork) testing methods are developed to link design
constraints for full-scale printed structures to material characteristics. Modular and conformally printed full-scale
wall prototypes are printed with a 6-axis robotic system. A Life Cycle Assessment of the prototypical earth
printing system is conducted, establishing a point of comparison to the climate impact of other construction
systems, including rammed earth, concrete masonry units, and 3D printed mortar. We demonstrate that printing
highly functional building elements with repeatable mechanical characteristics is possible using locally sourced
earth mixtures. By illustrating a range of reproducible material and geometric possibilities, we expand the design
space of additive earth and its applications.
1. Introduction
As global, rapid urbanization continues to grow, researchers and
practitioners in Architecture, Engineering, and Construction (AEC) are
exploring automation to accelerate the development of new housing and
infrastructure. At the same time, the construction industry is contrib-
uting 11% of global annual carbon emissions, a number projected to
grow over the next 50 years. Approximately 8% of this impact is a direct
effect of cement production for concrete, currently the world’s most
ubiquitous building material [45]. To address the twofold challenge of
rapid urbanization and its climate impact, novel systems must be
developed to pair low-carbon building materials with globally scalable
construction automation technologies. Building on existing earthen
construction systems, we situate novel, in-situ material processing and
testing methods for earth printing in a generalized framework for design
and construction with this emerging method of fabrication – 3D Printed
Earth (3DPE). We describe a method for Life Cycle Assessment for
earthen 3D printing and offer guidelines for scaling and applying 3DPE
based on the results of material and mechanical testing. By establishing a
design space for additive manufacturing with earth situated within
existing building practice through comparisons of material properties
and Life Cycle, we provide a new set of tools for architects and engineers
working to lower the embodied carbon of new construction.
Large Scale Additive Manufacturing is an area of growing research
and industry interest. There are now 3D-printed buildings worldwide,
some of which are inhabited [32]. All these structures engage a mix of
conventional construction practices with 3D-printed elements made
from concrete, plastic, clay, earth, or other experimental hybrid mate-
rials. The majority of 3D printed structures consist of walls made from
high cement content mortar (50–90% cement) where the 3D printed
portion of the building acts similarly to unconned masonry, serving as
formwork for conventional reinforced concrete columns and insulation
[33]. The oor and roof structure are constructed with conventional
techniques by skilled builders (cast concrete slab, timber framed). The
commonly stated advantages of 3D printing structures are:
1. Geometric freedom
2. Speed of construction
* Correspondence to: 77 Massachusetts Ave., Cambridge, MA 02139, USA.
E-mail address: curth@mit.edu (A. Curth).
Contents lists available at ScienceDirect
Construction and Building Materials
journal homepage: www.elsevier.com/locate/conbuildmat
https://doi.org/10.1016/j.conbuildmat.2024.135714
Received 25 November 2023; Received in revised form 3 March 2024; Accepted 4 March 2024
Construction and Building Materials 421 (2024) 135714
2
3. Reduced labor
4. Material savings [30].
The unique geometric exibility and efciency of 3D printing facil-
itate the construction of hierarchical structures that balance design with
performance.
This research investigates whether 3D printed earth could be a
realistic alternative to mortar that would effectively engage the pur-
ported advantages of additive manufacturing listed above while ful-
lling the same roles in the building system. The 3D-printed mortar
structures existing today represent a cautionary tale. Due to a lack of
3DP-specic building codes and models of 3D-printed mechanical
behavior, printed materials cannot easily be used for structural appli-
cations. In addition, cold joints formed between layers of rapid-setting
unreinforced mortar pose a challenging problem for material valida-
tion as the time between the deposition of layers varies for every new
geometry printed [5]. As a result, the exceptionally high strength
(60–120 MPa) accelerated mortar mixes that facilitate high-speed
printing and are typical in the industry are used for a formwork appli-
cation requiring an order of magnitude less strength. The carbon con-
sequences (though under-researched) are likely to outweigh any gains
made through material saving [50]. In addition, the thermal conduc-
tivity of concrete is high, meaning the inll structures common in 3D
printed concrete walls are signicant thermal bridges, reducing the
energy efciency of the printed buildings [48,49].
With minimal carbon impact, and the potential for passive thermal
performance, earth is a building material uniquely suited to 3D printing
directly recyclable, climate-friendly, and low-cost structures quickly in a
wide range of climates [11]. In addition to the general structural and
thermal benets of earth, the inherent re-resistant nature of the ma-
terial is worthy of considerable attention. According to the German
standard DIN 4102, Part 1, earth, even with ber additives, is not
combustible if its density is greater than 1700 kg/m3, making it
particularly relevant for the re-threatened context of Southern Cali-
fornia, where this research was conducted. Over the past decade, pro-
totypes have been 3D printed by IAAC researchers, WASP, and the
authors of this study, demonstrating the potential for performative
digitally fabricated earth-building systems [10,15,18]. However, limited
research has been published describing the material and mechanical
properties of 3D-printed earth and how it compares to other building
materials. In addition, few studies describe the fundamentals of design
and fabrication with earth in an additive manufacturing process or how
appropriate materials can be sourced. The continual challenge of this
research is validating the material and mechanical properties of vari-
able, locally sourced earth to meet common engineering standards for
widespread adoption without making the technology inaccessible
through prohibitively expensive laboratory testing methods. Through
literature review, we situate this study within the existing body of
research on earth as a building material, large-scale additive
manufacturing, and the climate impacts associated with both the ma-
terials and processes described. We then demonstrate foundational and
novel methods for Large-scale Earthen 3D printing, including material
processing, characterization for fabrication, and modular construction.
In addition, a preliminary Life Cycle Assessment (LCA) of earth printing
is presented based on the experimental setup used to produce the prints
referenced in this study, revealing marked differences in climate impact
between raw and stabilized earth printing and setting a carbon bench-
mark for future studies. By linking novel earth printing strategies
directly with their climate impact, a meaningful assessment of the
technology and its applications for sustainable design can be made.
1.1. Background and literature review
1.1.1. Earth as a sustainable construction material
Earth has been used as a building material for as long as humans have
made shelter. Currently, 1–3 billion people live and work in earthen
structures [38]. Earth-building technology has progressed in many di-
rections, from rammed earth (RE) to compressed earth blocks (CEB’s),
thermally and structurally performative, multistory structures found in
the Americas, Africa, Asia, and Europe (see Fig. 2). The demand for truly
low-carbon construction materials to combat climate change is rising in
the form of global Sustainable Development Goals and carbon reduction
programs in individual countries, states, and cities [51]. To meet these
demands, earth-building methods are increasingly being considered in
contemporary construction practice. Studies on the embodied carbon of
earthen building techniques indicate that unstabilized rammed earth
and compressed earth blocks can be used effectively in place of concrete
for 1/5 the carbon impact [29] as nonstructural building elements,
which offer a high degree of thermal performance. There are also ex-
amples of load-bearing earth systems up to 10 stories tall in Shibam,
Yemen, and studies describing unstabilized raw earth building elements
with compressive strengths from 1.5 to 4 MPa [21]. However, few
countries have adopted structural building codes for earth. Building
codes for earth in Australia and New Zealand are the exception and
include extensive guidance on earth wall building from a limit state
design approach for structures up to 6.5 m [42]. Nonstructural appli-
cations for earth also exist in building codes in the US state of New
Mexico, Zimbabwe, France, and Germany. As Loftness suggests, the
main challenge to the adoption of earthen building techniques in
building codes is a lack of engineering analysis and projects where
building performance is quantiable [34]. Beyond characterization as a
thermally and structurally performative building material, the critical
challenge to the widespread adoption of earth for contemporary con-
struction is the persistent colonialist perception of earth as a “primitive,”
or low quality building material [38]. 3D printing offers a decidedly
“modern” method of building with earth, which could address this
perception challenge and help validate the performance of other earth
building methods including RE and CEB construction.
1.1.2. Automating earthen construction
Building on traditions thousands of years in the making, rammed
earth is in the early stages of automation, with researchers engaging
computer-controlled pneumatic rammers and factory-style assembly
lines to produce structural earth modules [21,24]. A notable example is
the Kr¨
auterzentrum designed by Herzog and de Meuron and Martin
Rauche, where prefabricated rammed earth wall sections were assem-
bled on site. Similarly, earth block technology has become increasingly
automated through on-site mechanized compression systems that utilize
hydraulic presses to produce uniform, high density bricks rapidly.
Additional studies have explored hybrid digital fabrication strategies,
including daubed wicker [14], and Impact Printing, a process similar to
direct shaping or stacking of wet earth [39].
Earth has also been adapted to the much younger technology of
Large Scale Additive Manufacturing or Contour Crafting [30].
3D-printed earth buildings have been completed in Italy, Spain, the
United States, and Dubai [10,15,46]. Previous studies addressed mate-
rial processing and mechanical characteristics of 3D printed, unstabi-
lized earth [2,23,3,44]. These studies provide general reference points
for material formulations partially utilizing raw earth and demonstrate
average unconned compressive strength (USC) of 0.87 and 1.2 MPa
(see Appendix B). Ferretti et al. analyzed the material and structural
properties of the Italian 3D Printing Company WASP’s lime-stabilized
earth mixture, nding a USC of 2.32 MPa. More recently Faleschini
provided comprehensive insight into the effects of different cement and
lime stabilizers as well as natural ber types [17,18]. Asaf’s study offers
critical insight on fresh material properties of 3D printed engineered soil
mixtures and compressive/tensile test results of material, as well as a
comparison of 14 and 28 day strengths [2]. Bajpayee takes a deep look
at relevant rheology and soil chemistry for earth printing, however
unlike this study, the earthen materials used are heavily processed and
modied with geopolymers and plasticizing admixtures. Bajpayee also
indicates the importance of life cycle analysis and used dissolvable
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
3
Fig. 1. Recent examples of earthen additive manufacturing. Including a hybrid earth timber stair system produced by WASP/IAAC [1], WASP’s Tecla structure [41]
and Emerging Object’s “Casa Covida”( [46]).
Fig. 2. Earth construction covers much of our planet and ranges from large spanning structures [20] to 10-story buildings [43], to subterranean shelters [35]. Map of
earthen construction adapted from CRAterre [12].
Fig. 3. Robotic rammed earth forming systems with “active formwork” are under development [24], while systems like Echale’s hydraulic “Ecoblock” maker are
already widely used. (Images: ITE/TU Braunschweig, Echale).
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
4
geopolymers to promote circularity, though no information on the
embodied carbon of the whole process is given [3]. Ji compares the
properties of four potentially suitable soils for earth printing, comparing
their rheology[28]. Notably, only Gomaa’s tests include replicates (3),
while Perrot and Ferretti report the strength of singular mechanical
tests. While other examples of full-scale, load-bearing 3D earth-printed
structures exist, the mechanical properties of the materials used have
not been described ([10]; [46]). A key research gap exists in detailing
material processing and testing for load bearing earthen 3D printing
with minimally processed, locally sourced soil mixtures as opposed to
printing mixtures engineered from pure sand and clay feedstocks.
Perrot’s 2018 study focused on adding alginate to a soil mixture to
improve its green strength, facilitating the production of stable wall
sections taller than what could be produced in raw soil in a single
printing session. Gomaa’s 2021 study examined the potential for 3DPE
as a load-bearing material by simulating a set of case study wall ge-
ometries with strength values based on physical testing of three 3D
printed cob specimens with an average USC of 0.87 MPa. An additional
study by Gomaa examined the thermal performance of 3D printed cob
with various soils, nding thermal conductivity values (0.3–0.50 w/MK)
in line with other earthen materials relative to sample density and values
signicantly lower than equivalent 3D printed mortar geometries [22,
49] which also aligned with conventional cast concrete conductivity
values. Ferretti et al.’s single mechanical test is unique in that the ma-
terial is stabilized and amended earth (10% lime, 19% silica sand by
weight), and the test geometry is a hollow wall section with diagonal
inll. Each of these previous studies is compared to our results (Ap-
pendix B). Given the variable form factors and results of tests conducted
to date, research on 3D-printed earth is still in a nascent phase. A critical
goal of this study is to share insights on which material assessment
strategies are most accessible and relevant for standardized testing re-
gimes for the broad range of environments earthen buildings are suitable
for.
1.1.3. Material formulation / additives
Extensive studies have been conducted to establish the material
properties of soils suitable for raw earthen construction; however, as
Delgado found in a global survey, contemporary building codes around
the world include little guidance for building with earth. Those that do
(New Zealand, Australia, Germany, and the US State of New Mexico)
often suggest the addition of cement or other stabilizers to structural
earthen building elements to meet a minimum strength as specied by
the code [13]. Studies have shown that particle size distribution, ber,
and clay content can all play signicant roles in the compressive
strength of earthen mixes, allowing for improvements in strength and
durability without sacricing circularity [7]. Recently, Koutous
demonstrated that increasing the ber content of compressed earth can
improve compressive strength to a degree comparable to stabilization
with 1–2% cement [31]. Finding performative alternatives to cement
stabilization to produce structural earthen materials is necessary to
maintain their status as truly low-carbon, circular materials. When
cement is added to monolith building elements like CEB or rammed
earth, even in small quantities (1–5%), the overall cement content in a
structure can rival that of an entirely concrete building due to the thick
walls typically employed in earth systems [6]. In addition,
cement-stabilized earth cannot be directly recycled into new building
material. While our study does not aim to supply material character-
ization methods applicable to all forms of earth construction, the ap-
proaches used are consistent with the contemporary “best practices”
from the literature, including particle size distribution, soil type classi-
cation, and mechanical testing with the intention to produce code
compliant, unstabilized earthen building elements.
2. Methods
Material selection, processing, and characterization are critical
challenges for all earthen construction endeavors, given the inherently
variable qualities of locally sourced materials. These are also commonly
shared challenges for 3d printing systems, from the desktop to building
scale. Without consistent methods of material preparation and calibra-
tion, printing high-precision, performative geometry is untenable. As a
result, most 3D printing systems are run with highly engineered mate-
rials, from humidity-controlled thermoplastics to quick-curing mortars
formulated by industrial suppliers for use with proprietary extrusion
systems. At the architectural scale, this poses a major challenge for the
use of additive manufacturing in construction. Highly engineered ad-
mixtures are expensive, and their availability inconsistent. Reinforced
concrete is an excellent example of scalable construction technology
built on an extremely cheap, locally sourced set of materials that can be
calibrated in the eld with readily available additives like sand, gravel,
and water. In this study, we review existing relevant methods to make
earth a performative material for LSAM and present novel, on-site
testing methods to facilitate quick material assessment and adjust-
ment. We test print earth cylinders, novel methods for accurately
measuring their volume, and early-stage hydrostatic pressure tests for
the use of printed earth as formwork. We also share an approach for the
Life Cycle Analysis of 3D printed earth. For Large Scale Additive
Manufacturing, or digital fabrication more generally, to become widely
used in construction, the computational and mechanical systems
involved must be made compatible with local materials.
Fig. 4. (a) Perrot 2018 tested square sections of 3×2 layer linear extrusions, (b) Gomaa 2021 tested cylinders, (c) Ferretti 2022 tested a full-scale hollow wall section.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
5
2.1. Material sourcing
The rst step to any large-scale earthen construction project is
sourcing soil with a suitable mix of clay and sand. Often designated as
“sandy clay loam,” a soil made up of roughly 30% clay and 70% sand
offers a good balance of workability, low shrinkage rate, and cohesion.
Such soils can be found nearly anywhere, and simple strategies can be
employed to speed up the process of sourcing good material [13].
1. Find local earthen architecture; the best soil is often very nearby, or
the owners/builders may be willing to direct you. The color and
texture of the soil in local earth architecture may also offer a visual
guide as you examine local geology for quality building materials.
2. Seek out alluvial plains. Often, underneath a layer of rich soil lled
with organic material will be a layer of inorganic silt, sand, and clay
ideal for earth construction.
3. Areas in the foothills of mountains prone to erosion can provide a
good source as recent ooding events bring silty soils to the surface
of seasonal riverbeds and are recently deposited, so they have not
developed a topsoil lled with organic materials.
In the eld, an experienced earth builder can often assess the soil for
construction purposes simply by adding water and feeling the texture,
cohesion, and clay content by hand. Previous studies and construction
handbooks suggest a variety of eld tests to add consistency to these
somewhat subjective methods [36]. By creating a ball or rope of wet
earth, plasticity and cohesion can be assessed, giving an approximation
of clay content. When soil with an appropriate feel is found, increasingly
rigorous testing can commence.
For the tests and prototypes presented in this study, the above steps
were taken, beginning with a survey of two historic sites in Southern
California: the Mission Santa Barbara, constructed in 1786, and El Pre-
sidio de Santa Barbara, a Spanish fortress and housing complex con-
structed in 1782. While both adobe structures have been renovated
multiple times over the past 250 years, they contain bricks of consistent
color and texture. The material is familiar to local builders and referred
to as “Montecito mud,” a clay-rich, inorganic alluvial soil known for
causing drainage issues on construction sites and blocking roads in
annual winter landslides when it is brought down from local hills by
storm-driven erosion. As a result, all soil used in the following experi-
ments was sourced at no cost from construction sites within a 15-mile
radius of our 3D printing site. Local builders typically pay for the
removal of this highly desirable earthen building material from their
construction sites, making our application a mutually benecial op-
portunity for the community.
2.2. Material processing
To 3D print earth, local soil is rst processed to make it suitable for a
given project and pumping system. After soil with an appropriate ratio
of clay to sand is found through eld tests, it must be sifted to eliminate
large rocks or organic debris that are not pumpable. For all experiments
described below, the soil was air-dried and sifted through a ¼” /
6.35 mm screen constructed on a 1.25 ×1.25-meter frame positioned at
a 15-degree tilt (see Fig. 6). Moist soil can be sifted, however, clumping
reduces yield, so a dry mix is preferred. Industrial soil sifting equipment
such as conveyer fed shake tables can be used for higher throughput if
available. Following sifting, the soil is mixed with water and any
necessary additives, such as chopped straw or additional sand, to pro-
duce an extrudable and performative mix. After mixing, the earthen
material is ideally left to sit for 24 hours, fully hydrating the clay,
allowing it to reach a homogenous state, thus reducing variability in the
extrusion process caused by material inconsistency. Well-prepared ma-
terial is then loaded into the pumping system, and printing can
commence. Failed prints or materials that have been extruded, but are
not needed in the completed structure (formwork or temporary sup-
ports) may be directly cycled back into the pumping system so long as
they have not dried or been contaminated with large rocks or pieces of
concrete [9]. If part of a mix becomes contaminated, it may be resifted
before being introduced back into the extrusion system. Fully dried
prints may also be rehydrated to an appropriate water content and then
reused without a need to resift.
Fig. 5. Conceptual diagram of building continuously with minimally processed earth. 3D printing earth is a fundamentally circular building process.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
6
2.2.1. Fiber and additives
Earth for 3D printing can be amended with the same logic as rammed
earth, cob, or earth blocks (adobes). Various mix parameters may be
modied to improve mechanical behavior, or binders may be added to
alter compressive strength or durability. To reduce shrinkage, more sand
or aggregate is added. To improve cohesion and plasticity, more clay is
added. To improve even drying and overall strength, natural bers like
straw or rice husk may be added. Dening an optimal distribution of
ber sizing and quantity is a compelling area for innovation in the eld,
with recent developments indicating ber can provide equal stabiliza-
tion and improved ductility when compared to cement or lime additives
[31].
In the experiments detailed here, chopped wheat straw is the only
additive. The naturally hollow structure of straw which assists with even
drying, its rot-resistant properties, and wide availability have made
straw a common choice in traditional earth construction (barley straw is
particularly rot-resistant, though difcult to source near our site). In the
hot climate of southern California, rot is of less concern than
Fig. 6. Waste earth from construction sites can be sifted directly into a printable material if soil properties are favorable. Printed earth used as formwork can be
directly re-saturated back into a printable material (authors’ photos).
Fig. 7. Fully dried, printed earth can be rehydrated to a printable mixture within 24 hours.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
7
applications in wetter climates of Northern Europe or Asia. The straw is
roughly chopped to produce a gradient of lengths from 10 mm to
100 mm. Fibers that are shorter than the diameter of the extruder nozzle
maintain a stochastic orientation in extruded material, whereas bers
longer than the nozzle diameter are aligned inside the pumping system.
This anisotropic behavior can be leveraged to improve resistance to
lateral loads in applications such as formwork. Cracks typically occur
perpendicular to the printing direction, meaning extrusion-aligned, long
bers are ideal for reducing cracking. However, aligned bers often fail
to reach the print surface, reducing the efcacy of hollow bers like
straw for even drying and long-term moisture wicking, a typical benet
of hollow ber additives in CEB’s. A higher percentage of straw was
shown to be benecial for improved wet stability, uniform, crack-free
shrinkage (increased ductility), and enhanced compressive strength,
conrming recent studies by Koutous and Hilali in rammed earth. The
pump system used could not consistently tolerate an ‘ideal’ straw ratio of
0.75% by weight [31]. Initial experiments were conducted with 0.25%
straw content by weight. When increased to 0.50%, pressure at the
pump head ruptured the 25 mm ID, 150 PSI steel reinforced transfer
hose used in our standard system, puncturing the hose with pieces of
straw. With a hose rated to 250 PSI, we did not encounter this issue
when running a mix with 0.50% straw content. There are many syn-
thetic and natural bers with the potential to enhance the material
properties of earth for additive manufacturing; our study is designed to
serve as a baseline for future research on this topic.
2.2.2. Mixing
After earth is sifted, water is added to achieve a pumpable viscosity.
For this study, a ratio of 1:5 water to sifted soil was used. Mixing was
performed in a standard ve cubic foot concrete mixer. To accurately
add water, the starting water content by mass of the sifted soil must be
determined by weighing a sample before and after oven drying (ASTM
D2216). The soil used in our experiments had a starting water content of
10%+-4%. If available, standard viscosity modiers can be added to
reduce the water-to-soil ratio. However, such additives pose a risk of
producing uncharacterized/unpredictable behaviors such as separation
and clumping in your mix if your soil chemistry reacts poorly with the
VMA (Viscosity Modifying Agent). In addition, they may be carbon and
cost-intensive, reducing the efcacy and scalability of earth printing as a
sustainable building strategy. Minimizing water in the mix is essential to
limit shrinkage and resultant undesirable cracking behavior during
drying. Conversely, the water content used in 3D printing earth mixes is
necessarily higher than that of CEB or rammed earth to facilitate
pumping. An advantage of higher water content is that extruded mate-
rial typically has a higher density than either CEB or rammed earth. A
balance must be found between the shrinkage rate and the desired nal
material density. Further studies are required to quantify the relation-
ship between water content and aeration.
Water is added to the mixer prior to the incremental addition of soil
and ber. Unlike concrete mixing, earth mixing is ideally not a “just in
time” process. The farther ahead an earth printing mix can be prepared,
the more homogenous the extrusion will be. As in ceramics
manufacturing practice, mixes can be left for days, weeks, or even
months, “aging” them to achieve a truly homogenous mixture. Clay,
especially unprocessed clay in locally harvested soils, takes time to
saturate fully. This can be observed when a fresh, un-aged mix of earth is
loaded into a pumping system, and the extrusion appears to progres-
sively thicken over the course of a day, eventually reaching an
unpumpable viscosity, overheating the pump, clogging the transfer
hose, and generally disrupting work. Managing mix consistency on-site
through well-timed batch mixing is critical for consistent material ow.
If space and time allow, all material can be prepared well in advance and
left to fully saturate for at least one week to mitigate inconsistent
extrusion. Future research could include the integration of soil satura-
tion tests for on-site material assessment.
2.2.3. Pumping
The authors’ system utilizes a portable progressive cavity pump that
accepts aggregates up to 6.35 mm (1/4”) (IMER Small 50). Many pump
systems in the concrete, water, and food industries can potentially be
implemented in earthen material delivery to the end effector of an LSAM
system. The current industry standard in C3DP (Concrete 3D Printing), i.
e., mortar printing, is the progressive cavity pump. Pump size is driven
by the desired ow rate, hose run (head pressure), and aggregate size. In
this study, one of the smaller, easily portable solutions was chosen for
exibility and cost. Progressive cavity pumps include wear parts (the
rotor), which may need frequent replacement depending on how abra-
sive the aggregate in each soil mix is.
2.3. 3D printing system
This study used a 2014 IRC5 ABB 4600 2.55 m industrial robot arm
with a 40 kg payload. The robotic arm was sourced from a car factory
assembly line in Detroit, MI, and was refurbished to maintain its original
factory tolerance of 0.01 mm per meter. This model, in particular, was
chosen for its balance of machine weight, ease of transportation and
repositioning (machine weight: 500 kg) to reach (2.55 m radius), and
payload (40 kg), which is adequate for 3D printing end effectors and
nominal transfer hose loads. For the tests conducted, the arm was
mounted on a 60 cm pedestal and tted with 30 cm - 100 cm extensions
at the end effector to maximize the build area of the machine. Such
extensions can be simply fabricated using sheet steel or available steel
pipe products. As evidenced by the proliferation of LSAM systems
globally, there are many 3–7 axis systems and system designs available
to choose from for any given application. As a relatively low-cost system
(~$30,000 for both arm and pump), we found this setup to be ideal for
prototyping 1:1 scale architectural elements in 3D printed earth. Limi-
tations on print speed are driven by mechanical limits, material delivery
limits, or material curing/drying limits. In earthen additive
manufacturing material delivery is typically the key factor for speed as
minimizing water content drives buildability, but puts greater strain on
the pump system than a low viscosity mortar mix that is accelerated at
the end effector.
2.4. Material evaluation
The Food and Agriculture Organization of the UN, as well as Cra-
Terra, provide simple eld tests for the initial evaluation of soils. Such
tests can be paired with geotechnical, structural, and chemical labora-
tory tests to create a comprehensive soil evaluation. For most earth ar-
chitecture applications, however, physical on-site soil analysis is
adequate to begin work. These tests can include wetting the soil and
feeling its texture, drop tests, and ribbon tests to assess clay content and
cohesiveness [8,25]. These tests are common in practice and well
documented in the literature. In this research, eld tests, as well as
laboratory tests, were conducted to determine soil type, particle size
distribution, moisture content, specic gravity, and Unconned
Compression Strength (USC). In addition, a soils report was obtained
from the contractor whose waste soil we were utilizing (Appendix B).
Examples of critical laboratory tests for the validation of an earthen
building product can be found in Delgado 2005.
2.4.1. Jar test
A small sample of dry soil is added to a jar of water. Results are easier
to interpret if the ratio of soil to water is at least 1:2 by volume. The jar of
soil and water is shaken vigorously and left to settle for 24–48 hours.
Three layers of grain size and type will be visible: clay at the top, then
silt, and sand at the bottom. By simply comparing the volume of clay and
silt to sand, decisions can be made about the viability of the raw soil for
earth construction. It is not uncommon to amend the soil with another
local source of either clay or sand to achieve a desirable ratio after
conducting this simple test. Typical mixes suitable for construction are
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
8
approximately 70% sand and 30% silt and clay, with as little organic
material in the mix as possible. In the jar test, organics can often be seen
oating in the top of the water column. In the case of our material,
“Programmable Mud” (PM), a relatively smooth or, “well graded,”
particle size distribution results in poorly dened separation between
materials in the jar test. In this study a capped, graduated cylinder is
used in place of a jar to improve the accuracy of results. Ultimately, this
proved to be a positive indicator for key printable soil properties.
2.4.2. Specic gravity
Specic Gravity of the raw soil used in this research was measured in
accordance with ASTM D854.
2.4.3. Particle size analysis
Particle Size Analysis of the raw soil used in this research was con-
ducted in accordance with ASTM D442.
2.4.4. Shrinkage
The ultimate dry shrinkage of an earthen printing mix is a critical
factor for constraining design. To monitor shrinkage, we developed a
plastic 3d printed square stamp 30 mm x 30 mm (see Fig. 8). By
stamping a print directly after extrusion, its shrinkage could be directly
measured parallel and perpendicular to extrusion direction as it dried.
2.4.5. UCS (unconned compressive strength)
To determine the UCS of 3d printed earth, samples were tested in
accordance with ASTM D2166, a standard for measuring the compres-
sive strength of cohesive soils. 10 cm x 20 cm (4in x 8in) solid cylinders
were printed with a 20 mm layer height. Each cylinder took approxi-
mately two minutes to print. Samples were left to dry at approximately
20 C for four weeks. Height and diameter measurements were taken to
determine the volume of the samples after they were printed and dried.
In addition to standard top, middle, and base diameter measurements,
the difference in diameter between the peak and trough of the extruded
layers was also measured and factored into the volume calculation of
each cylinder. To account for variation in diameters across peak and
trough, LIDAR 3D scanning was used to generate digital mesh models of
cylinders so the volume could be digitally computed and the diameter
cross-checked.
UCS tests were conducted on a Baldwin 300 Hydraulic testing ma-
chine at a rate of 10 kN/sec. UCS was calculated using load and position.
No displacement sensors were placed on the surface of the cylinders,
given their uneven texture. A platen with a ball seat was used for the
USC tests, given the samples’ expected low strength and variability.
Efforts were made to ensure parallel faces on the top and bottom of the
cylinder by sanding the contact surfaces smooth with a set of parallel
guides prior to weighing, measuring, and load testing. Samples were
prepared following the ASTM C617 standard for sulfur mortar bonded
caps. The method of sulfur capping was selected for its adherence to mud
samples and its effectiveness in creating parallel surfaces on non-
standard cylindrical geometries. The capping enabled uniform distri-
bution of loads for axial loading perpendicular to the capped surfaces,
allowing closer adherence to ASTM C39 specications for planeness and
perpendicularity. Parallel samples were tested with and without sulfur
capping as a control for discrepancies in the testing process (See Fig. 13).
Plaster capping was explored but bonded poorly to the earth samples.
2.4.6. Bridging test
A test was designed to establish the bridging limits of printed earth
material. Three horizontal layers of extrusion were printed across a se-
ries of unsupported spans. These spans stretched from ’column’ to
’column,’ each spaced at increasing distance intervals (see Fig. 14). The
vertical deection in the horizontal path over the gaps between columns
was then assessed. A simple map was produced for each layer, where “no
go” is a section of extrusion that sinks completely below the plane of the
extruder nozzle. An intermediary category was added to indicate areas
where the bearing surface area of the extrusion thinned due to down-
ward deformation but still maintained some at surface in the plane of
the nozzle. This intermediary level of deformation is notable because, in
all cases tested, the subsequent layer would be a complete “go.” A "go" is
dened as an extrusion that maintains a uniform extruded surface from
column to column.
2.4.7. Double cone test
A test was devised to assess the maximum print overhang for an earth
mix. Two cones are printed, one sloping inward from bottom to top and
one sloping outward from bottom to top. By producing pairs of cones at
different slop angles ranging from 5 to 30 degrees, the capacity for
cantilevered printing in idealized minimum tension (inward facing
cone) and maximum tension (outward facing dimension) can be quickly
determined. Each double cone test took approximately 10 minutes to
print.
2.4.8. Hydrostatic pressure test
To test the performance of 3D printed earth as formwork for either
cast-in-place or prefabricated reinforced concrete construction, a
method for hydrostatic pressure testing was developed. Hollow cylin-
ders measuring 100 mm in interior diameter and 260 mm in height
(height limited to minimize slump during fabrication) were printed and
lled with wet concrete at different drying stages. Rigid plastic tubing
was used to extend the height of hydrostatic pressure tests conducted on
cylinders that had dried to strengths, necessitating greater pressure to
induce failure. Each hydrostatic pressure test cylinder took approxi-
mately 5 minutes to print.
2.5. Life Cycle Assessment
While raw earth is directly recyclable, the material processing,
fabrication, and eventual demolition of a 3D-printed earth structure
does have a climate impact. Given that no 3D-printed earth buildings
Fig. 8. Calibrated stamps are used to asses shrinkage throughout the drying process. Stamps are measured throughout the drying process over the course of
several days.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
9
have been inhabited or demolished and subsequently recycled to date,
an LCA including operational energy, or end-of-life is not possible at this
time. However, we can quantify a cradle-to-cradle life cycle assessment
of a test print referencing existing approaches to the Life Cycle Assess-
ment of earth architecture [16]. Processing energy costs, water usage,
and fabrication energy impacts can be correlated to the volume of
extruded material, a value directly extracted from the digital model of
the 3D print toolpath. The result is an embodied carbon cost per meter of
extrusion for our particular LSAM setup. We can then calculate the
carbon cost of a functional unit of 3D-printed earth. We consider a per kg
model and present a volumetric comparison with related building ma-
terials (see Fig. 19). The 2019 ICE database and the 2023 Carbon
Leadership Forum’s North American Material Baseline report are
referenced to produce a reasonable comparison with common building
materials.
Given that 3D printed concrete is not included in these databases,
individual LCA studies and reviews of 3DPC LCA studies were referenced
[47,50] for comparison across printed materials. The embodied carbon
required to produce and service the ABB arm and Pump were not
factored into the LCA as they operate on the order of 10’s of thousands of
hours over their lifetime and require minimal maintenance. The resul-
tant carbon impact is extremely small relative to the per kg of material
impact. All assumptions about embodied energy are conservative,
assuming machines are running at maximum power and are sourced
from supplier power specications. Future studies could include direct
power measurements during the printing process.
3. Results
Based on the methods outlined in section three, the experimental
Fig. 9. Categorizing an earthen material’s sand, silt, clay ratio quickly allows for assessment of its suitability for earthen 3D printing. A material does not need to
match a perfect ratio of particle sizes so long as it is within the region of feasibility dened by the literature for compressed earth blocks and rammed earth.
Fig. 10. The particle size distribution of “Programmable Mud” indicates a ratio of particle sizes conducive to a high packing density. A visual depiction of grain size
in useful for assessing material quality in the eld.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
10
results of this research are structured to provide reproducible guidance
for future studies and construction in 3D printed earth. Performative
material properties are described (4.1.1–4.1.5), then tests for estab-
lishing a design space in a given earthen material are demonstrated
(4.1.6–4.1.8). Additional tests illustrate the potential for earthen 3d
printing as formwork for reinforced concrete (4.1.9) and as a method for
building modular earthen wall systems (4.2). Finally, a Life Cycle
Assessment methodology is presented within the context of existing
building materials and large-scale additive manufacturing processes
(4.3).
Fig. 11. A well-graded soil results in higher compressive strengths. Note the material used in this research is a well-graded soil similar in particle size distribution to
an engineered CMU block mix.
Fig. 12. Capping cylinder tests prior to loading resulted in more consistent results and higher strengths, likely a product of more uniform loading.
Fig. 13. Fractured cylinder tests. Capped USC tests indicated an average compressive strength of 3.4 MPa, a value comparable to rammed earth with 1–2% cement
added (Appendix B).
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
11
3.1. Material properties
3.1.1. Soil type
As evidenced by earth-building traditions and studies from around
the world, a range of soil types are viable for building. Well-graded soil
for earthen construction is typically 20–30% clay, 50–60% sand, and no
more than 10% coarse particles (gravel or larger aggregates) to ensure
good compaction. Fig. 9 illustrates the sand/silt/clay ratio of the Sandy
Clay Loam used in this research, as well as a conservative range of soil
types that could be viable for earth printing based on experience in the
eld (+- 10% in each axis of the Clay, Silt, Sand, ratio) [13]. This range
is a exible design space in which builders can tune their material
relative to available local resources. As the mix is adjusted, different
hydration times, adjustments to water content, and extrusion speeds
may need to be explored to maintain a stable and consistent output with
minimal shrinkage.
3.1.2. Specic gravity
A specic gravity of 2.13 was found, indicating there is either
organic or porous matter in our soil. Organic-rich soil for agricultural
applications is typically under 2, while pure sand is between 2.65 and
2.85 [20]. Notably the estimated specic gravity of the raw excavated
soil as reported in the source construction site’s soils report was 2.44
(Appendix B).
3.1.3. Particle size analysis
The particle size distribution of the earthen material used in this
research (PM) can be compared to that of previous earthen materials
studies and engineered concrete particle size distributions. As prior
geotechnical studies have indicated, a uniform PSD improves material
packing density, making for a denser earthen matrix with higher
compressive strength [20]. The comparison in (Fig. 11, Table 3) indi-
cated a consistent result with the literature: “well-graded” materials
have higher density and mechanical performance [2]. Particle Size
distribution is a critical factor in assessing the quality of raw earthen
material sources on site. Future studies could investigate the relation-
ship and tradeoffs between packing density, strength, and drying time as
well as reasonable adjustments to be made to a local material to improve
its PSD.
3.1.4. Shrinkage
Over the course of test prints ranging in size and geometry from small
cylinders to two-meter wall sections a shrinkage rate of 4% +-2% was
observed. Tests of the standard mix at water contents by weight of 10%,
15%, and 20% indicated greater water content is correlated to higher
shrinkage rates, a result conrmed by the literature. Further small-scale
mix tests indicate a 2–3% shrinkage rate can be achieved in a standard
PM mix at 15% water content, which has been amended with 5%
decomposed granite sand. However, full-scale printing tests were not
conducted with this amended material as assessing minimally processed
earth was the key goal of the study.
3.1.5. Unconned compressive strength
The primary focus of compression testing was to assess the structural
performance of a novel method of fabrication with a standard earthen
mix, not to test a novel material. With strengths ranging from 2.2 to
3.7 MPa, 3D printed earthen samples from this experiment meet or
exceed the highest strength raw earth construction in the literature,
indicating the fabrication process may have a critical effect on strength.
The dry density of the 3D-printed earth also exceeds that of all examples
found in previous studies. As noted by other researchers, dry density is
closely related to compressive strength in earth-based materials [26,40].
The marked difference in strength and spread between uncapped and
sulfur-capped samples clearly indicated capping is a critical step for
consistent results. Another potential factor in the high strength of our
tests is the extrusion-aligned distribution of straw bers. Rings of straw
ber may resist hoop stresses, which could be a critical mode of buckling
failure in this material. Further studies, including rammed earth or CEB
samples in matching material and controls without ber, would be
necessary to characterize the relationship between raw earth, ber
Fig. 14. A standardized bridging test allows designers to assess critical material constraints quickly.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
12
properties, and the effects of the 3D printer pump system on strength.
This structural test of a novel material can guide future studies meth-
odology as well as provide a basis for more accurate simulation of the
printed earth’s mechanical behavior.
3.1.6. Bridging test
The bridge test is primarily a quick method to diagram the design
space of a given material, however specic quantitative metrics can also
be extracted. A ratio between spannable distance and adjacent, hori-
zontal toolpath spacing is evident. By the third layer, the material used
in this test could cleanly span a gap of the same width as the extrusion,
so long as there was full contact with an adjacent horizontal toolpath.
Bridging gaps greater than extrusion width, up to double extrusion
width, was only possible with adjacent toolpath overlap over 20%.
These metrics are critical for designing complex 3D geometry where
consistent wall thickness may be a factor for even drying. The ability to
bridge, and design for bridging parameters, opens opportunities for
material saving and greater geometric complexity in earth-based print-
ing strategies. A similar test could be directly applied to other earthen
and cementitious 3D printing materials, including quick-setting mortars
that are common in industry.
3.1.7. Double cone test
The maximum inward cone angle for the PM mix was 25 degrees, the
maximum outward cone angle was 20 degrees. Cones are an idealized
print condition, elastic failure is more likely at lower angles where
inherently unstable geometry like rectilinear overhanging wall section
are printed, or where excess material is added by layer transitions. To
produce as evenly extruded a test as possible we positioned the layer
change seam between the two cones. Like the bridging test, the double
cone test serves as a quick assessment tool for understanding what is
geometrically possible with a given local earth material.
3.1.8. Failure modes
The common failure modes in 3D printed earth observed during the
production of tests, wall prototypes, and previous large-scale experi-
ments are:
1. elastic buckling during fabrication, and
2. cracking during drying
3. inconsistent extrusion
Buckling is caused by a combination of unstable geometry and self-
weight exceeding the bearing capacity of the freshly printed material.
Some degree of elastic deformation from self-weight is typical when
printing earth and can be measured after printing by comparing nominal
layer height from the base of the print to the top.
Cracking can be caused by three forces: self-weight during fabrica-
tion, when loaded beyond its dry compressive strength, and differential
shrinkage during drying (most common – see Fig. 16). Producing data
for each mode of failure allows for the creation of an accurately con-
strained design space.
3.1.9. Hydrostatic pressure
Preliminary hydrostatic pressure tests were conducted on 3D printed
earth of cylindrical and rectilinear geometry at varying stages of drying
to understand the material’s potential as formwork. Buckling failure was
achieved in a cylindrical test (260 mm tall, 100 mm interior diameter)
loaded with a Quikcrete 27 MPa High Strength Concrete mixed to
50–75 mm slump (ASTM C143). A freshly printed earth cylinder failed
at a hydrostatic pressure of 0.106 MPa, when it had been lled to a
height of 200 mm with concrete (see Fig. 17). However, we were unable
to produce signicant deformation in test geometries dried at room
temperature (6 hours, 12 hours, 24hrs).
3.2. Case study: modular 3D printed earth wall
Once the material behavior of a given local earth mix is established,
practical considerations of design performance come to the forefront. To
demonstrate the viability of 3D printed earth as a modern construction
Fig. 15. Double cone tests. Top - 15 degree slope, Bottom - 30 degree slope.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
13
Fig. 16. Common failure modes - a. surface tearing from heterogeneous mix, b. cracking due to differential drying, c. hose rupture at the pump head, d. buckling
under self-weight during printing.
Fig. 17. Wet concrete is added incrementally to the freshly printed cylinder until failure from hydrostatic pressure. Buckling occurs after hoop stresses overcome the
tensile limits of the extruded cylinder’s walls.
Fig. 18. Post-tensionable earthen wall modules can be mass-manufactured and transported to site.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
14
Fig. 19. A spatial consideration of performance vs carbon illustrates the relationship between new technologies and old. Notably, 3D printed concrete is sized at the
minimum width printable with the mix referenced and is therefore over strength.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
15
technology, we designed a prototypical modular wall system that can be
printed in a factory, an onsite printing facility, or in situ on a con-
struction site. A planar or conformal starting condition is considered.
Building on previous experimental structures, the design balances sta-
bility during printing, voids for insulation, and contact points for inte-
rior nishes. A layer height of 20 mm was used in all tests. A post
tensioning plate and rod system facilitates easy craning of the nished
module as well as in situ stability as a compressed element where
stresses from lifting elements into place are evenly distributed across the
print surface. Post tensioning low strength materials is an area of
developing research [4]. The resulting structures illustrate the range of
geometric exibility feasible with earth printing technology as well as
the critical ability to link site specic, adaptive printing with pre-
fabricated modular elements. The inherently versatile nature of additive
manufacturing can make complex, performative geometries in local
materials like earth which were previously inaccessible because of
monolithic manufacturing processes.
3.3. Life Cycle Assessment
We found the key contributing factor to climate impact in the 3D
printed earth process is the energy required to pump material to the end
effector. While still a low carbon cost when compared to conventional
construction material systems, this pumping impact could be mitigated
by direct feed hopper systems [23] or more efcient pumps that utilize
energy from a renewable grid mix. The results indicate that similar to
rammed earth or compressed earth blocks, the embodied carbon of
printed earth as a building material is less than 1/5 of a 20 MPa concrete
[29] and 1/50 of a typical 3D printed mortar.
While the functions of earth and concrete, in terms of structural and
thermal performance, are markedly different, it is critical to note that
currently, the vast majority of 3D printed mortar structures are lost
formwork for conventional reinforced concrete structural systems. As
such, the potential for 3D printed earth as minimal, recyclable, re-
resistant, and optimized formwork for both reinforced concrete and
conventional insulation materials places it in an entirely new low-
carbon design space. There is great potential to implement the mate-
rial and energy-saving strategies outlined in a growing body of academic
research [19,27] through the geometric design exibility of additive
manufacturing, particularly when paired with the low carbon and
thermal comfort benets of earth.
4. Discussion
This research indicates that readily accessible earthen materials,
often considered waste on construction sites, can be directly applied to
the construction of building elements that are structurally performative,
geometrically complex, and have low carbon impact. The methods
employed to produce and load test printed cylinders may be applicable
to other materials and a step towards more standardized approaches for
the LSAM eld. The mechanical tests detailed in this research demon-
strate that 3D printed earth can have load-bearing properties exceeding
previous 3DPE, rammed earth, and CEB/adobe studies. A minimal
technical assessment of this soil was conducted before producing rela-
tively high-performing tests, indicating the vast opportunity for future
studies examining the qualities of locally available soils, particularly
those used in nearby traditional construction. Analysis of the rheology of
the 3DPE material indicates that particle size distribution, density, soil
type, and ber additives all play critical roles in the selection of a
performative local earthen material for 3D printed construction. Fabri-
cation of super-meter prototypes further illustrates the broad possibil-
ities for earthen 3D printing in design indicated the technology could be
safely applied to the construction of housing and infrastructure in
appropriate seismic zones.
The key limitations of the earth printing process are environmental.
Print speed is entirely driven by drying speed as wet earth needs to gain
enough strength through drying to bear additional layers of material. In
areas with low humidity and high temperatures or wind, printing may
progress quickly, in wetter, colder environs, the pace of construction
may be greatly reduced. Available materials and the comparatively high
water content relative to CEB or rammed earth are additional potentially
critical limitations driven by local resource availability. Higher water
content and aggregate limitations of available pumping systems also
result in higher shrinkage than other earth-building techniques, which
must be accounted for in material calibration and the design of printed
building elements.
5. Conclusions
Basic research on a range of locally sourced, printed earth mixes
would expand our understanding of the generalized viability of this
emerging technology. The work presented in this paper is only a rst
step towards building a rational and reproducible approach to engaging
earth additive manufacturing as a part of modern construction practice.
Future work could include the assessment of anisotropic tensile behavior
produced by ber alignment in printed earth and a comparative study of
earth-printed structures’ resistance to erosion and water intrusion
compared with CEB or rammed earth. Studies of printed earth’s strength
throughout the drying process and its durability over time in different
climates would also be of critical interest for larger scale industry
adoption. On-site hydrostatic pressure testing methodologies are
another area to explore to understand printed earth’s potential as a
material for formwork. From a mechanical perspective, the development
of larger material conveyance systems capable of handling lower water
contents and larger aggregates could further advance the scalability of
the technology. While these systems exist, they are typically outside the
budget of academic research, requiring industry involvement to expand
this area of development. As the push for sustainable and cost-efcient
construction continues to grow in response to climate change and
rapid urbanization, earthen additive manufacturing has the potential to
create safe, performative, and scalable housing.
CRediT authorship contribution statement
Lawrence Sass: Writing – review & editing, Supervision. Caitlin
Mueller: Writing – review & editing, Supervision, Methodology. An-
gelica Castro-Salazar: Visualization, Methodology, Investigation,
Formal analysis. Natalie Pearl: Writing – review & editing, Visualiza-
tion, Investigation. Alexander Curth: Writing – review & editing,
Writing – original draft, Visualization, Validation, Supervision, Soft-
ware, Resources, Project administration, Methodology, Investigation,
Funding acquisition, Formal analysis, Data curation, Conceptualization.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data Availability
Data will be made available on request.
Acknowledgments
Funding was provided by the MIT Programmable Mud Initiative.
Thank you to Oliver Moldow for assisting in the fabrication of the
3DP earth bridging test, Stephen Rudolph for assistance load testing
samples, and LYNDA Labs for providing fabrication space and support.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
16
Appendix A. : Experimental parameters
Table 1
Mix design for "Programmable Mud"
Mix design - 1:5:0.025 % by weight
Dry Soil sifted to ¼” 1
Water 0.2
Straw 0.25–0.050
Table 2
Print Parameters of the experimental setup
Print Parameter Specication
Nozzle Diameter 25 mm
Extrusion width 50 mm
Print Speed 50–300 mm/second
Appendix B. : Results data
Table 3
Earthen and cementitious mix properties.
% Passing
Sieve Size (mm) Programmable
Mud
Rammed Earth
(532)
Rammed Earth
(433)
Rammed Earth
(424)
CEB 6% wt Hydraulic
Lime
CEB - Laterite
Clay
CMU Stone Dust
0.063 1.63% 20% 20% 40% 34% 14% 3%
0.125 4.91% 22% 22% 41% 42% 18% 5%
0.25 18.29% 30% 35% 47% 50% 25% 8%
0.5 36.03% 47% 52% 62% 64% 35% 10%
2 86.57% 69% 61% 71% 83% 63% 36%
4 99.57% 65% 66% 76% 88% 90% 88%
6.3 100.00% 73% 73% 82% 93% 99% 98%
10 100.00% 100% 100% 97% 95% 100% 100%
14 100.00% 100% 100% 100% 100% 100% 100%
Material Properties
Density (g/cm
3
) 1.93 2.08 2.18 2.16 1.95–1.99 2.78 1.5–2.0
Compressive Strength
(MPa)
3.4 0.9 1.4 1.4 9 2.81 20
Soil Type (as specied by
source)
sandy clay loam silty clay silty clay silty clay clay silty clay sand
Source
This study Hall, Djerbib
2004
Hall, Djerbib
2004
Hall, Djerbib
2004
Teixeira, et. al 2020 Seick, et. al
2018
Sucksun, et. al
2014
Table 4
Soil plasticity metrics, as indicated by the soils report
from the construction site where the waste earth used
for this study was sourced.
Plasticity Index (ASTM D4318)
Liquid Limit 44
Plastic Limit 15
Plasticity Index 29
Table 5
Soil expansion metrics, as indicated by the soils report
from the construction site where the waste earth used for
this study was sourced.
Expansion Index (ASTM D4829)
Expansion Index 44
Expansion Potential Low
Initial Saturation 50%
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
17
Table 6
Soil density metrics, as indicated by the soils report from the construction site
where the waste earth used for this study was sourced.
Maximum Density (ASTM D1557)
Mold Diameter 4 in. (10.16 cm)
Weight of rammer 10.00 lbs (4.54 kg)
No. of Layers 5
No. of Blows 25
Est. Specic Gravity for 100% Saturation Curve 2.44
Maximum dry density 113.5 pcf
Optimum water content 12.4%
Table 7
Compression test results
Sample Density (g/cm^3) Ultimate Strength (MPa) E - steepest slope (Pa) UCS Sd UCS CV
A-1 2.06 2.30 185.89
A-2 2.10 3.08 179.90
A-3 2.03 2.18 165.41
A-4 2.04 2.36 161.57
A series average 2.06 2.48 173.19 0.41 0.16
B-1 1.86 3.69 165.31
B-2 1.95 3.61 135.48
B-3 1.93 3.29 165.36
B-4 1.98 2.98 225.57
B-5 1.92 3.51 292.84
B series average 1.93 3.42 196.91 .29 .08
Table 8
Life Cycle Assessment Inputs and Results
Material Source Density (kg/m3) UCS (kN/M^2) Embodied Carbon (kgCO2/kg)
Rammed Earth (RE) ICE 2019, [37] 1850 2000 0.023
Compressed Earth Block (CEB) ICE 2019,
Morel 2005
1950 4000 0.010
Stabilized CEB (5% cement) ICE 2019, Morel 2005 1950 10000 0.060
Concrete Masonry Unit (10 MPa) ICE 2019, CCMPA 2000 10000 0.073
Fired Clay Brick ICE 2019 1900 17000 0.230
3D Printed Concrete Roux 2022, Papachristoforou 2018 2000 120000 0.208
3D Printed Earth (WASP) Ferreti 2022 1815 2300 0.077
3D Printed Earth (PM) This study 2000 3400 0.006
References
[1] Alice, IAAC and WASP: new 3D printing strategies towards the realisation of load-
bearing earthen structures, IAAC (Blog) May 29 (2019) 2019. 〈https://iaac.net/iaa
c-wasp-new-3d-printing-strategies-towards-realisation-load-bearing-earthen-stru
ctures/〉.
[2] Ofer Asaf, Arnon Bentur, Pavel Larianovsky, Aaron Sprecher, From soil to printed
structures: a systematic approach to designing clay-based materials for 3D printing
in construction and architecture, Constr. Build. Mater. 408 (December) (2023)
133783, https://doi.org/10.1016/j.conbuildmat.2023.133783.
[3] Aayushi Bajpayee, Mehdi Farahbakhsh, Umme Zakira, Aditi Pandey, Lena Abu
Ennab, Zoa Rybkowski, Manish Dixit, et al., In situ resource utilization and
reconguration of soils into construction materials for the additive manufacturing
of buildings, Front. Mater. 7 (March) (2020) 52, https://doi.org/10.3389/
fmats.2020.00052.
[4] Martin Bechthold, Zach Seibold, Saurabh Mhatre, Post-tensioned ceramic
structures: design, analysis and prototyping, Archit. Struct. Constr. February
(2022) 2, https://doi.org/10.1007/s44150-022-00025-0.
[5] Freek Bos, C. Menna, Mauricio Pradena, Eric Kreiger, W.R. Silva, A.U. Rehman,
Daniel Weger, et al., The realities of additively manufactured concrete structures in
practice, Cem. Concr. Res. 156 (June) (2022) 106746, https://doi.org/10.1016/j.
cemconres.2022.106746.
[6] Steve Burroughs, Strength of compacted earth: linking soil properties to stabilizers,
Build. Res. Inf. 34 (1) (2006) 55–65, https://doi.org/10.1080/
09613210500279612.
[7] Steve Burroughs, Soil property criteria for rammed earth stabilization, J. Mater.
Civ. Eng. 20 (3) (2008) 264–273, https://doi.org/10.1061/(ASCE)0899-1561
(2008)20:3(264).
[8] “Compressed Earth Blocks Standards (CDI - CRATerre-EAG, 1998, 144 p.), n.d.
Accessed October 10, 2023. https://www.nzdl.org/cgi-bin/library?e=d-00000-
00—off-0cdl–00-0——0-10-0—0—0direct-10—4—————0–0 l–11-en-50—20-
about—00-0-1-00-0-0-11-1-0utfZz-8-
10&cl=CL4.30&d=HASH01979938ef89e979ddfb736b>=2.
[9] Curth, Alexander. Pending. 3D Printed Earth Formwork for Reinforced Concrete
Construction. US Patent Ofce 63/507,725, led June 12, 2023.
[10] Curth, Alexander, Barrak Darweesh, Logman Arja, and Ronald Rael. 2020,
Advances in 3D Printed Earth Architecture: On-Site Prototyping with Local
Materials, In BE-AM | Building Environment Additive Manufacturing, 105–10.
Darmstadt.
[11] Curth, Alexander, Eduardo Gascon, Lawrence Sass, Leslie Norford, and Caitlin T.
Mueller. 2023, Additive Energy: 3D Printing Thermally Performative Building
Elements with Low Carbon Earthen Materials, Research Gate. https://doi.org/
10.13140/RG.2.2.14662.65600.
[12] David Gandreau and Leticia Delboy, World Heritage Inventory of Earthen
Architecture, CRATerre-ENSAG, France, 2012. 〈https://unesdoc.unesco.org/ark:/4
8223/pf0000217020〉.
[13] M. Delgado, Ignacio Ca˜
nas, The selection of soils for unstabilised earth building: a
normative review, Constr. Build. Mater. 21 (February) (2007) 237–251, https://
doi.org/10.1016/j.conbuildmat.2005.08.006.
[14] D¨
orstelmann, Moritz. 2023, Digital Wicker, Text. Author. KIT. August 14, 2023.
https://ddf.ieb.kit.edu/353.php.
[15] Dubor, Alexandre, Jean-Baptiste Izard, Edouard Cabay, Aldo Sollazzo, Areti
Markopoulou, and Mariola Rodriguez. 2019, On-Site Robotics for Sustainable
Construction, In Robotic Fabrication in Architecture, Art and Design 2018, edited by
Jan Willmann, Philippe Block, Marco Hutter, Kendra Byrne, and Tim Schork,
390–401. Cham: Springer International Publishing. https://doi.org/10.1007/978-
3-319-92294-2_30.
[16] Pierre Est`
eve-Bourrel, Christopher Beckett, Remo Pedreschi, F. Bosche, Jean-
Claude Morel, Rabia Charef, Guillaume Habert, Developing an integrated BIM/LCA
framework to assess the sustainability of using earthen architecture, IOP Conf. Ser.:
Earth Environ. Sci. 1078 (September) (2022) 012100, https://doi.org/10.1088/
1755-1315/1078/1/012100.
[17] Flora Faleschini, Daniel Trento, Maryam Masoomi, Carlo Pellegrino,
Mariano Angelo Zanini, Sustainable mixes for 3D printing of earth-based
constructions, Constr. Build. Mater. 398 (September) (2023) 132496, https://doi.
org/10.1016/j.conbuildmat.2023.132496.
A. Curth et al.
Construction and Building Materials 421 (2024) 135714
18
[18] Elena Ferretti, Massimo Moretti, Alberto Chiusoli, Lapo Naldoni, Francesco De
Fabritiis, Massimo Vison`
a, Mechanical properties of a 3D-printed wall segment
made with an earthen mixture, Materials 15 (2) (2022) 438, https://doi.org/
10.3390/ma15020438.
[19] Gasc´
on Alvarez, Eduardo, Caitlin Mueller, and Leslie Norford. 2021. Dynamic
Thermal Performance of Structurally Optimized Concrete Floor Slabs. https://doi.org/
10.26868/25222708.2021.31052.
[20] Minke Gernot, Building With Earth: Design and Technology of a Sustainable
Architecture, Birkh¨
auser, Berlin, 2006.
[21] Giada Giuffrida, Rosa Caponetto, Massimo Cuomo, An overview on contemporary
rammed earth buildings: technological advances in production, construction and
material characterization, IOP Conf. Ser.: Earth Environ. Sci. 296 (July) (2019)
012018, https://doi.org/10.1088/1755-1315/296/1/012018.
[22] Mohamed Gomaa, Wassim Jabi, Alejandro Veliz Reyes, Steve Goodhew,
Jim Carfrae, Thermal performance exploration of 3D printed cob, Archit. Sci. Rev.
62 (April) (2019), https://doi.org/10.1080/00038628.2019.1606776.
[23] Mohamed Gomaa, Wassim Jabi, Alejandro Veliz Reyes, Veronica Soebarto, 3D
printing system for earth-based construction: case study of cob, Autom. Constr. 124
(April) (2021), https://doi.org/10.1016/j.autcon.2021.103577.
[24] Mohamed Gomaa, Sascha Schade, Ding Wen Bao, Yi. Min Xie, Automation in
rammed earth construction for industry 4.0: precedent work, current progress and
future prospect, J. Clean. Prod. 398 (April) (2023) 136569, https://doi.org/
10.1016/j.jclepro.2023.136569.
[25] Harmonized World Soil Databasev1.2 | FAO SOILS PORTAL | Food and Agriculture
Organization of the United Nations, n.d. Accessed November 2, 2021. https://
www.fao.org/soils-portal/data-hub/soil-maps-and-databases/harmonized-world-
soil-database-v12/en/.
[26] Houben, Hugo, and Hubert Guillard. 1989, Earth Construction: A Comprehensive
Guide, In. Rugby, Warwickshire, United Kingdom: Practical Action Publishing.
https://doi.org/10.3362/9781780444826.
[27] Mohamed A. Ismail, Paul L. Mayencourt, Caitlin T. Mueller, Shaped beams:
unlocking new geometry for efcient structures (September), Archit., Struct.
Constr. (2021), https://doi.org/10.1007/s44150-021-00003-y.
[28] Yameng Ji, Philippe Poullain, Nordine Leklou, The selection and design of earthen
materials for 3D printing, Constr. Build. Mater. 404 (November) (2023) 133114,
https://doi.org/10.1016/j.conbuildmat.2023.133114.
[29] Jones, Craig, and Geoffrey Hammond. 2019, Inventory of Carbon & Energy (ICE),
http://www.circularecology.com/ice-database-faqs.html.
[30] Behrokh Khoshnevis, Automated Construction by Contour Crafting—Related
Robotics and Information Technologies, in: Automation in Construction, 13, The
best of ISARC, 2004, pp. 5–19, https://doi.org/10.1016/j.autcon.2003.08.012,
2002.
[31] A. Koutous, E. Hilali, Reinforcing rammed earth with plant bers: a case study,
Case Stud. Constr. Mater. 14 (June) (2021) e00514, https://doi.org/10.1016/j.
cscm.2021.e00514.
[32] Nathalie Labonnote, Anders Rønnquist, Bendik Manum, Petra Rüther, Additive
construction: state of the art, challenges and opportunities, Autom. Constr. 72
(September) (2016), https://doi.org/10.1016/j.autcon.2016.08.026.
[33] Junli Liu, Vuong Nguyen-Van, Biranchi Panda, Kate Fox, Anton du Plessis,
Phuong Tran, Additive manufacturing of sustainable construction materials and
form-nding structures: a review on recent progresses, 3D Print. Addit. Manuf. 9
(1) (2022) 12–34, https://doi.org/10.1089/3dp.2020.0331.
[34] Loftness, Vivian, Kent Harries, Erica Hameen, and Lola Ben Alon. 2017, Integrating
Earthen Building Materials and Methods into Mainstream Housing Projects
throughout Design, Construction, and Commissioning Stages, In.
[35] Luigi Rosselli Architects. 2015, The Great Wall of WA, Luigi Rosselli Architects
(blog). June 19, 2015. 〈https://luigirosselli.com/residential/the-great-wall-of-wa〉.
[36] Lyle A. Wolfskill, et. al. 1969, Handbook for Building Homes of Earth, Washington
D.C.: Department of Housing and Urban Development, Division of International
Affairs.
[37] Vasilios Maniatidis, Peter Walker, Structural capacity of rammed earth in
compression, J. Mater. Civ. Eng. 20 (3) (2008) 230–238, https://doi.org/10.1061/
(ASCE)0899-1561(2008)20:3(230).
[38] Alastair T.M. Marsh, Yask Kulshreshtha, The state of earthen housing worldwide:
how development affects attitudes and adoption, Build. Res. Inf. 50 (5) (2022)
485–501, https://doi.org/10.1080/09613218.2021.1953369.
[39] Coralie Ming, Ammar Mirjan, Jesús Medina Ib´
a˜
nez, Fabio Gramazio,
Matthias Kohler, Impact Printing, 3D Print. Addit. Manuf. 9 (3) (2022) 203–211,
https://doi.org/10.1089/3dp.2021.0068.
[40] Jean-Claude Morel, Abalo Pkla, Peter Walker, Compressive strength testing of
compressed earth blocks, Constr. Build. Mater. 21 (2) (2007) 303–309, https://doi.
org/10.1016/j.conbuildmat.2005.08.021.
[41] Massimo Moretti, WASP in the Edge of 3D Printing (Digital Innovations in
Architecture, Engineering and Construction), in: 3D Printing for Construction with
Alternative Materials, Springer International Publishing, Cham, 2023, pp. 57–65,
https://doi.org/10.1007/978-3-031-09319-7_3.
[42] “NZS 4297:2020, 2020. 〈https://www.standards.govt.nz/shop/nzs-42972020/〉.
[43] “Old Walled City of Shibam - UNESCO World Heritage Centre, n.d. Accessed June
1, 2022. 〈https://whc.unesco.org/en/list/192/〉.
[44] A. Perrot, D. Rangeard, E. Courteille, 3D printing of earth-based materials:
processing aspects, Constr. Build. Mater. 172 (May) (2018) 670–676, https://doi.
org/10.1016/j.conbuildmat.2018.04.017.
[45] Martin R¨
ock, Marcella Ruschi Mendes Saade, Maria Balouktsi, Freja
Nygaard Rasmussen, Harpa Birgisdottir, Rolf Frischknecht, Guillaume Habert,
Thomas Lützkendorf, Alexander Passer, Embodied GHG emissions of buildings –
the hidden challenge for effective climate change mitigation, Appl. Energy 258
(January) (2020) 114107, https://doi.org/10.1016/j.apenergy.2019.114107.
[46] Ronald Rael, Virginia San Fratello, Alexander Curth, and Logman Arja. 2020, Casa
Covida: Mud Frontiers III, ACADIA 2020 | Distributed Proximities.
[47] Roux, Charlotte, Kateryna Kuzmenko, N. Roussel, Romain Mesnil, and Ad´
elaïde
Feraille. 2022, Life Cycle Assessment of a Concrete 3D Printing Process, The
International Journal of Life Cycle Assessment, November. https://doi.org/10.1007
/s11367-022-02111-3.
[48] Jingting Sun, Jianzhuang Xiao, Zhengrong Li, Xiwen Feng, Experimental study on
the thermal performance of a 3D printed concrete prototype building, Energy
Build. 241 (June) (2021) 110965, https://doi.org/10.1016/j.
enbuild.2021.110965.
[49] Thadshajini Suntharalingam, Irindu Upasiri, Perampalam Gatheeshgar,
Keerthan Poologanathan, Brabha Nagaratnam, Paulo Santos,
Heshachanaa Rajanayagam, Energy performance of 3D-printed concrete walls: a
numerical study, Buildings 11 (10) (2021) 432, https://doi.org/10.3390/
buildings11100432.
[50] Matheus Pimentel Tinoco, ´
Erica Martinho de Mendonça, Letícia Ikeda
Castrillon Fernandez, Lucas Rosse Caldas, Oscar Aurelio Mendoza Reales, Romildo
Dias Toledo Filho, Life cycle assessment (LCA) and environmental sustainability of
cementitious materials for 3D concrete printing: a systematic literature review,
J. Build. Eng. 52 (July) (2022) 104456, https://doi.org/10.1016/j.
jobe.2022.104456.
[51] A.A. Wieser, M. Scherz, S. Maier, A. Passer, H. Kreiner, Implementation of
Sustainable Development Goals in Construction Industry - a Systemic
Consideration of Synergies and Trade-Offs, in: In IOP Conference Series: Earth and
Environmental Science, 323, Place of Publication: Graz, Austria. Country of
Publication: UK.: IOP Publishing, 2019, https://doi.org/10.1088/1755-1315/323/
1/012177.
A. Curth et al.
