Fracture Mechanics of Concrete and Concrete Structures -

Recent Advances in Fracture Mechanics of Concrete - B. H. Oh, et al.(eds)

ⓒ 2010 Korea Concrete Institute, Seoul, ISBN 978-89-5708-180-8

The toughness of imperial roman concrete

P. Brune & R. Perucchio
University of Rochester, Rochester, New York, USA
A.R. Ingraffea
Cornell University, Ithaca, New York, USA
M.D. Jackson
Northern Arizona University, Flagstaff, Arizona, USA

ABSTRACT: The concrete composites used to realize the monumental structures of Imperial Rome are re-
markable engineering materials. While the endurance of intact constructions such as the Pantheon evinces the
concretes’ durability, such durability mostly serves to preserve the mechanical properties, which are responsi-
ble both for the monuments’ original creation and continued survival. Despite their prominent role in the en-
gineering achievements of the empire, these mechanical properties – particularly in tension and fracture –
have not been comprehensively assessed. We first review the mechanical properties obtained through various
experimental programs conducted on both authentic ancient composite core samples and their components,
summarizing the major findings and outlining the remaining gaps in knowledge. We then qualitatively discuss
the fracture of Roman concrete within the context of our own testing program, which will test both re-
fabricated and authentic materials, with the aim of characterizing the fracture behavior that has contributed to
the preservation of a significant component of engineering heritage.

1 INTRODUCTION fragmented brick and volcanic rock coarse aggregate

Between 60-160CE, Imperial Roman engineers
honed their usage of concrete to create spanned mo-
numental structures with designs that would violate
present-day Civil Engineering building safety codes
(ACI 318-08 2008). And yet a surprising number of
these buildings are still extant today, some in excel-
lent states of preservation and covered by their orig-
inal unreinforced concrete vaults. Scholarly investi-
gations tend to emphasize the architectural
significance of the monuments instead of exploring
the structural considerations invoked by the creation
and survival of these daring constructions. Those
rare studies that analyze structural behavior have
been constrained by limited knowledge of the me-
chanical behaviors of the constituent concretes. In-
deed, not a single analysis has incorporated the frac-
ture mechanics of Roman concrete as a quasi-brittle
material. Such analyses are critical for both preserv-
ing deteriorating structures and understanding the
endurance of those still intact, as well as studying
the ancient design conventions responsible for their
conception and construction.
Fracture mechanics has not previously been ap-
plied to Roman concretes, mainly because little is
known of their physical behavior as cementitious
composites. The conglomeratic fabric contains
of decimeter-scale dimensions bonded by a poz-
zolanic mortar, based on altered volcanic ash ini-
tially mixed with hydrated lime. The mortar has a
relatively low compressive strength as compared
with Portland cement mortars. The published me-
chanical testing data is sparse, and provides only
scattered compressive strength values that are not
accompanied by full load-displacement curves. Va-
riability in mortar and coarse aggregate composi-
tions further reduces the applicability of these re-
sults, particularly as they pertain to fracture of the
concretes and the buildings realized therein. All told,
describing the fracture behavior of these highly het-
erogeneous composites with widely varying con-
stituents presents remarkable challenges. However,
its characterization is vital to accurately assessing
the safety of surviving structures, and understanding
their extraordinary durability in response to a com-
bination of differential settling on weakly consoli-
dated bedrock and seismic ground motions over their
nearly 2000-year life spans.
We begin with an exploration of published and
unpublished mechanical test data for both the con-
glomeratic concretes and their assorted constituents.
Examination of results for authentic historic speci-
mens, modern laboratory-fabricated re-productions,
and raw geologic materials provides initial insights
into the potential application of modern concrete = − D (of
The structural Jfabric h, TRoman
) ∇h concretes can be
fracture mechanics to describe the fracture behavior described on several length scales. On the structural
of the ancient cementitious composites. From this scale, the material The occupies large volumes
proportionality – 3 the D(h,T)
coefficient
review, we formulate several hypotheses about the Great Hall encompasses
moisture permeability and it is a ofnonlinea
an excess of 3000m
fracture mechanisms of the concretes. We include a concrete – as a heterogeneous
of the relativecomposite
humidity hcontinuum.
and temperature
preliminary description of our proposed testing On the meso level, & Najjar 1972). The moisture deci-
a pozzolanic mortar bonds mass balanc
program, which will employ a novel experimental meter-sized coarse thataggregates
the variation (caementa
in time).ofVarious
the water mas
configuration to measure the fracture properties of materials were used as caementa
volume , with(water
of concrete builders oftenw) be eq
content
laboratory-reproductions of an Imperial pozzolanic vertically gradingdivergence
the aggregatesof thebymoisture
mass density
flux J to
mortar, before culminating in the fracture testing of reduce the self-weight of upper sections of struc-
a significant volume of authentic ancient core sam- tures, especially in vaults.
∂
ples from the Great Hall of the Markets of Trajan − w
= ∇•J
∂
(c.110CE). t

The water content w can be expressed a

of the evaporable water we (capillary wa
2 DESCRIPTION OF ROMAN CONCRETES vapor, and adsorbed water) and the non-e
(chemically bound) water wn (Mil
The meaningful discussion of Roman concretes is Pantazopoulo & Mills 1995). It is reas
necessarily intertwined with the structures it real- assume that the evaporable water is a fu
ized. Accordingly, to provide context for a review of relative humidity, h, degree of hydration
the mechanical properties of ancient concretes, we degree of silica fume reaction, αs, i.e. we=w
first examine the Great Hall in terms of its concretes = age-dependent sorption/desorption
to view Imperial Roman monumental concretes (Norling Mjonell 1997). Under this assum
through the lens of an exemplary structure. The by substituting Equation 1 into Equati
Great Hall is an appropriate choice because, through obtains
the generosity of those charged with its care, we
have been able to meaningfully study the mechanical ∂w ∂h ∂w ∂w
behavior and material composition of its structure in Figure 2. Computational e model
− solid Dhthe
+ ∇ • (of ∇h)Great
= e α& show-
c +
Hall, e α& + w
unprecedented depth (Jackson et al. 2009, Brune & ∂h ∂t
ing the three structural-scale materials. ∂α
c ∂α
s
s

Perucchio, in prep.).
When approaching the Great Hall, one sees main- The wall and where
vault concretes
∂we/∂h is the of the
slopeGreat
of theHall
sorption/
ly the brick facing that clads the concrete nucleus of (Fig. 2) feature atisotherm
least two(alsodistinctcalled
aggregate mix- capac
moisture
standard Imperial Age wall construction (Fig. 1). tures (Jackson et al. 2009). The
governing caementa
equation of the 3)
(Equation wall
must be
The considerable dimensions of monumental build- concrete include byfragmented
appropriate bricks (~1600kg/cu.m)
boundary and initial conditi
ings, combined with the tenacious bond between and two tuffs: the The compact and between
relation relativelythedurable
amount of e
core and cladding, consign the facing to curing and Tufo Lionato (~1700kg/cu.m)
water and relative theand porousisand
humidity called ‘‘
weather protection functions while ensuring that the weakly durable isotherm”
Tufo Giallo della Via with
if measured Tiberina
increasing
conglomeratic core acts as a structural skeleton. Al- (~1500kg/cu.m). humidity
The lighter andvault concrete isotherm”
‘‘desorption contains in th
though hidden from view, the concretes at the center almost exclusively Tufo Giallo della Via Tiberina.
case. Neglecting their difference (Xi et al.
of Imperial monuments were hardly afterthoughts. The wall mortarthe wasfollowing,
produced‘‘sorption
by combining
isotherm”hy-will be
Indeed, these are remarkably complex materials, in- drated lime withreference
Pozzolaneto Rosse altered volcanic
both sorption and desorption c
corporating a diverse mixture of constituents from ash aggregate, mainly
By the in small
way, gravel-
if thetohysteresis
sand-sizedof the
Rome’s rich geologic surroundings to form a versa- scoriae, and veryisotherm
small quantities
would be taken into sand-
of ground, account, two
tile and durable building material. sized Tufo Lionato aggregate.
relation, evaporableThe binding
water matrix of humi
vs relative
the wall mortarbecontains alumina- and alkali-
used according to the sign of the varia
cement gels and relativity
strätlingite.humidity.
In contrast, The theshape
vault of the
mortar contains isotherm
notable for amounts (~33 volume%)
HPC is influenced by many p
of light grey pumice (~800kg/cu.m),
especially those thatainfluence
smaller quan-
extent and
tity (~15% volume%)
chemical reactions and, in Rosse
of heavier Pozzolane turn, determ
(~1700kg/cu.m), structure
and a very andsmall
porequantity (<5 vol- (water-
size distribution
ume%) of ground Tufo Lionato. The cements
ratio, cement chemical composition, are SF
as yet unknown.curing
Takentime together, the two concrete
and method, temperature, mix
formulations evince
etc.).Roman
In the builders’
literature sophisticated
various formulatio
and intentional deployment
found to describe of available materials
the sorption isotherm
towards the controlling of self-weights. But how
concrete (Xi et al. 1994). However, in th did
Figure 1. Photo of brick-faced concrete wall (bottom left) with the diverse material
paper combinations translate into
the semi-empirical me- pro
expression
schematic showing concrete nucleus. A typical facing brick is chanical properties?
Norling Mjornell (1997) is adopted b
about 15cm wide.

Proceedings of FraMCoS-7, May 23-28, 2010

 − D (h, T )∇h EXPERIMENTAL TESTING OF (1)
3J =PREVIOUS explicitly
gate; relativeaccounts
proportions for the wereevolution
not recorded.of hydration
Uniax-
ROMAN CONCRETES & COMPONENTS reaction
ial compressionand SFtests content.
recorded Thisbothsorption isotherm
the compressive
The proportionality coefficient D(h,T) is called reads (mean=3.60MPa, stdev=1.96MPa) and elas-
strength
The
moisturenumerous confections
permeability and itofisImperial
a nonlinear Roman con-
function tic modulus (mean=2.9GPa, stdev=1.8GPa) of the
crete
of theare marked
relative by widely
humidity h anddivergent
temperature aggregate con-
T (Bažant samples. Several complete ⎡ stress-strain curves ⎤ were
stituents. Due to the general
& Najjar 1972). The moisture mass balance requires unavailability of au- also obtained.
we (htests Also,
α c α s ) = G (α c α s )⎢ −
significantly,
⎢ modulus
1 ⎥ of rup-
+ for
thentic specimens and the challenge
that the variation in time of the water mass per unit of laboratory re- ture ,
measured the bending(g tensile
,
1
, 1
∞
α − α )h ⎥ strength
⎥
fabrication, no comprehensive
volume of concrete (water content w) be equal to the experimental testing two Hadrian’s Villa samples ⎢
⎣ e (0.68,
10
c 0.78
1 c ⎦ MPa). (4)
of these various
divergence of thematerials
moisture has flux been
J executed to ro- However, no additional information concerning
∞ − α )h ⎤ curves spe-
bustly characterize their overall mechanical behav- cimen size or Kcomplete ⎡ (g α
load-displacement
(α α ) e ⎢
10
c c − ⎥
ior.∂ The experimental programs to date have focused was provided for the c bending
s ⎢ tests.
1
, 1

⎥
(2)
1
− = ∇on• Jcompressive strength. These findings con-
largely
w
In addition, samples from ⎣ the Basilica of ⎦Maxen-
∂
firm the expected variation in performance among
t
tius included cores from a large section of the col-
the The
different
waterformulations,
content w canwhich encourages
be expressed as thean sum
ex- where main
lapsed the first
vault,term two (gelof theisotherm)
barrel vault represents
base walls, the
amination of the mechanics of
of the evaporable water we (capillary water, water the constituent aggre- physically
and bound (adsorbed)
the foundation. water and the
Specific information second
about the
gate
vapor, components.
and adsorbed water) and the non-evaporable term is(capillary
cores limited toisotherm)their geometry represents the capillary
and bulk density.
(chemically bound) water wn (Mills 1966, water. This expression is
The aggregate constituents were not recorded. valid only for low content
All
Pantazopoulo & Millsroman 1995).concretes
It is reasonable to of SF. The
samples were coefficient
cylindersGwith 1 represents the amountand
15cm diameters of
3.1 Testing of ancient water per
heights unit volume
between 30 andheld 38cm, in the
exceptgel pores
for at 100%
three pris-
assume that the evaporable water is a function of relativeblocks
humidity, and it can be expressed (Norling
For
relativesampleshumidity, cored hfrom, degree Ancient structures,αcLam-
of hydration, , and matic of 35x35cm 2
cross section and 45cm
precht
degree(1984) of silicapublished
fume reaction,the firstαsmodern
, i.e. we=mechanical
we(h,αc,αs) Mjornell 1997) as
height. The samples were extracted along an axis
test results in the form
= age-dependent of compressive strengths,
sorption/desorption isotherm perpendicular to the stratification of the aggregate
with a single measurement
(Norling Mjonell 1997). Under of thethis elastic
assumption modulus and layers.
G (α c αThe volumetric
c s proportion of mortar in the
(18GPa). Cores were obtained
by substituting Equation 1 into Equation 2 one from a variety of concrete
1
, )= k α c + k α s
swas vgsaidc to fall vg sbetween 40-60%, with (5) the
sites,
obtainsstructures (from a roughly 250-year range), additional remark that the lowest proportions of
and locations within the structures. Lava, tuff (type mortar
where kappear
c
vg andin ksvgthearezones of theparameters.
material structure whereFrom the
unspecified),
∂w ∂h
basalt, sandstone,
∂w
quartz,
∂w
and slate are building
maximum amount of water per unit volume that Un-
process was executed “most carefully.” can
all
− mentioned
e as coarse
+ ∇ • ( D ∇h ) = aggregate.
e α&c + Diverse
e α&s + w&composi- (3) iaxial
fill all tests
pores measured
(both capillary the compressive
pores and gelstrength pores), one of
∂ h ∂ t h
tions of mortar were alsoc observed, ∂ α ∂ α with ndifferent nine samplesK(mean=4.6MPa,
can calculate 1 as one obtains stdev=1.4MPa) and
types of limestone – either dolomitic or pure – iden- s the elastic modulus of three (mean=2.7GPa, stdev=
tified
whereas∂w the quicklime source and pozzolana (prove-
e/∂h is the slope of the sorption/desorption
0.9GPa). No complete load-displacement ⎡
⎜ g α − α ⎟h ⎥
⎛ ∞ ⎞ curves
⎤
nance,
isotherm color, (also alteration
called facies,
moisture and maximum
capacity). grain
The were reported w − (Giavarini et
α s + α s −G ⎢ −e al. 2006).
⎢
⎝
10
c 1 c ⎠
size were not noted)(Equation
only identified forbecertain sam- As a whole, previous
0.188
c 0.22
smechanical 1

testing
⎥
of
⎥ an-
governing equation 3) must completed ⎦ (6)
0 1
⎢

ples.
by coarseNo information
appropriate boundary aboutandthe relative
initial percentages
conditions. cient
K (α c αconcrete
,
s )= cores provides scattered
⎣
compressive
of aggregate and mortar was provided. Lim- strengths over statistically ⎜ g α insignificant
⎛
c − α c ⎟⎠h −
∞ ⎞
and highly
1

The relation between thedimensions

amount ofofevaporable
10

ited information about the the pris- e ⎝

varying populations. Specimen sizes were generally
1
1

water
matic and
test samplesrelative humidity
was included, is called ‘‘adsorption
usually in relativity
the form onlyThe twomaterial
or threeparameters
times largerkc than thes caementa, the
isotherm”
of an edge if measured
length with increasing
(mean=6.7cm), except for two cy- largest heterogeneity, further vg and k vg and g1 can
restricting
humiditysamples
lindrical and ‘‘desorptionof 15cm isotherm”
diameter. in thewide
The opposite
vari- be calibrated
bility of the by fittingresults
strength experimental
to datatherelevant
structural-scale
applica- to
be-
case. Neglecting
ability in composition their difference
and provenance(Xi et al. of 1994),
the sam-in free (evaporable)
havior. Furthermore, wateraggregatecontentconstituents
in concretewereat
the
ples following,
is evident ‘‘sorption isotherm” will be used with various ages (Di Luzio & Cusatis 2009b).
reference
measured bothinsorption
tocompressive the considerable
and desorption
strengths, with
scatter
a
of the
conditions.
mean of
seldom catalogued with sufficient rigor. Altogether,
it seems that the composite nature of the ancient ma-
By the way,
12.9MPa if the hysteresis
(stdev=5.5MPa) over 52 ofsamples.
the moisture terial and the source
2.2 Temperature of its vulnerability in structures
evolution
isotherm
Samuelli would be taken
Ferretti (1996) into account,
oversaw two different
mechanical test- –Note
mainly tensile fracture – werethenotchemical
sufficiently well
relation,
ing evaporable
of Ancient Roman water vs
concrete relative humidity, must that, at early age, since reactions
be used
the Port according
of Trajan at sign ofcores
to Fiumicino
the and
obtainedoffrom
the Hadrian’s
variation the
Villa
understood
associated
goals. Only
to inform
with
two cement
small
thehydration
samples
experimental
were and
used SFprograms’
to reaction
measure
relativity
in Tivoli. for humidity.
TheHPC Fiumicino The shape
cores were of the
taken sorption
from the are exothermic,
tensile strength, the and temperature
no field
measurement is not
of uniform
fracture
isotherm
foundations of a is influenced
warehouse and bymeasured
many parameters,
15cm in for non-adiabatic
properties systems even if the environmental
especially
diameter those that influence extent and rate of the temperaturehasisbeen published.
constant. Heat conduction can be
chemical and
alternating
betweenand,
reactions
strata of
25.5 inandturn,
brick and
28cmdetermine
tuff
in height, with
(unspecified) pore described in concrete, at least for temperature not
structure
coarse and
aggregate. pore size distribution
The relative (water-to-cement
proportions ofcontent,
brick, exceeding
3.2 Component 100°C testing(Bažant & Kaplan 1996), by
ratio,and
tuff, cement
mortar chemical
were composition,
measured by area SFfrom maps Fourier’s
The law, which
variability evident reads in the composite responses
curing
of the In time
exterior and method,
core surfaces temperature,
as (35, mix additives,
20, 45), (2.5,
etc.).
62.5), and the(19, literature
16, 65). various
The formulations
samples from can 35,
Hadrian’s be asserts
q = − λ ∇T
behaviors
the importance of the component mechanical
to achieving a more robust estimation(7) of
found
Villa to describe the sorption isotherm of normal
concreteconsisted
10x10x14cm (Xi3 etandal.one of1994).
two However,
that was
prismaticin blocks
11x11x22cm the 3present
. Both
of the overall composite response. Several testing pro-
where have
grams q ismade the valuable
heat flux, T is theby absolute
contributions measur-
paper were
blocks the semi-empirical
taken from the expression
collapsed proposed
vaulting of theby temperature, and is the heat
ing the mechanical properties of mortar and coarse
λ conductivity; in this
Sala a Tre Esedre and contained only brick aggre-it
Norling Mjornell (1997) is adopted because aggregate constituents of the ancient concretes.
Proceedings of FraMCoS-7, May 23-28, 2010
3.2.1 Ancient roman bricks
Samuelli Ferretti (1996) tested ancient bricks, com-
monly used in fragmented form as caementa. Bricks
of widely varying provenance and quality, generally
taken from wall facings, were sawed to create 29
sample sets. From these, prismatic specimens meas-
uring 15x15x30mm3 were subjected to compression
(mean=17MPa, stdev=5.9MPa) and direct tension
tests (mean=3.33MPa, stdev=1.25MPa). The elastic
modulus (mean=13.4GPa, stdev=4.7GPa) was also
measured along with a ductility ratio, defined to be
the quotient of the strain at ultimate failure (after
J = − D (material
indicate quite scattered h, T )∇h characteristics and
rock strengths.
The proportionality coefficient D(h,T)
moisture permeability and it is a nonlinea
of the relative humidity h and temperature
& Najjar 1972). The moisture mass balanc
that the variation in time of the water mas
volume of concrete (water content w) be eq
divergence of the moisture flux J
− ∂ = ∇•J
w

∂
softening) and the strain corresponding to the peak t

stress. This ratio averaged 2.26 (stdev=0.35); a few
complete stress-strain curves are described as “char-
acteristic of a fragile material”.
3.2.2 Ancient roman mortars
Samuelli Ferretti (1996) created pozzolanic mortar
samples using hydrated lime and sieved Pozzolane
Rosse combined in Vitruvian proportions (1:3, by
volume). Only ash particles smaller than 2mm were
used; in contrast, the mortars in the Great Hall in-
clude Pozzolane Rosse scoriae up to 1.5cm. The
amount of Roman tap-water used was specified ac-
cording to a volumetric ratio with lime of (1.39:1); it
was not stated how much water was used to initially
hydrate the lime, making a water to cement ratio un-
The water content w can be expressed a
of the evaporable water we (capillary wa
vapor, and adsorbed water) and the non-e
(chemically bound) water wn (Mil
Pantazopoulo & Mills 1995). It is reas
assume that the evaporable water is a fu
relative humidity, h, degree of hydration
degree of silica fume reaction, αs, i.e. we=w
= age-dependent sorption/desorption
(Norling Mjonell 1997). Under this assum
by substituting Equation 1 into Equati
Figure 3. Strength data for Tufo Lionato and Tufo Giallo della
obtains
Via Tiberina. Each column corresponds to a different quarry.
Large markers represent the mean for each type of test result.
∂w 2007)
After De Casa (1999, ∂h and Jackson ∂we(2005). ∂we
− e + ∇ • ( D ∇h ) = α&c + α&s + w
et al. et al.

available. Prismatic beams of 40x40x160mm were 3

∂h ∂t h ∂α ∂α
cast, de-molded after an unspecified amount of time, c
The variation in results (Fig. 3), likely in part due s
and cured in a lime-water solution. After curing at 7, to dissimilar experimental setups, occurs not just be-
28, 90, 180, and 360 days, modulus of rupture tests tween quarries but where is the slope
also∂wfore/∂hdifferent of theand
locations sorption/
were followed by compression tests on the broken isotherm
stratigraphic levels in a single (also called
quarry. The moisture
scattered capac
halves. The results indicate a marked reduction in governing
strengths of the two equation
tuffs reflect their (Equation
heterogeneous,3) must be
flexural (30%) and compressive strength (10%) be- pyroclastic fabrics, which are composed of variable conditi
by appropriate boundary and initial
tween the 90-day and 180/360-day samples. This proportions of vitric, Thelithic,
relationand between the amount of e
crystal fragments
suggests a complex hardening process, perhaps due bound by zeolite water and relative
(and calcite) cements.humidity is called ‘‘
Furthermore,
to chemical phase transitions in the pozzolanic ce- isotherm” if measured
weathering of pumice glass in some tuff specimens with increasing
ments during curing. For the 360-day samples, a produces clay mineral that weakens the cohesion in th
humidity and ‘‘desorption isotherm”
mean flexural strength of 0.95MPa (stdev=0.10MPa) of the tuff, therebycase. Neglecting
reducing their difference
mechanical strength and (Xi et al.
and mean compressive strength of 12.07MPa durability (Jacksontheetfollowing,
al. 2005). ‘‘sorption isotherm” will be
(stdev=1.02MPa) were measured. The mean elastic reference can
Tuff coarse aggregates to bothformsorption
50 volume%and desorption
of c
modulus, measured during compression tests, was By the way, if the
the concrete fabric of monumental walls and vaults. hysteresis of the
3GPa (stdev=0.1GPa). isotherm wouldand
Therefore the characterization be inclusion
taken intoofaccount,
the two
relation,
mechanical behaviors evaporable
of the Roman water vs relative humi
tuff lithologies
3.2.3 Roman volcanic tuffs be used according
is central to an accurate compositetomaterial the signmodel
of the varia
Various tuffs from the Roman region were fre- describing the fracture of Imperial concretes. The of the
relativity humidity. The shape
quently used as caementa in monumental construc- isothermtuffformechanical
widespread variability HPC is influenced by many p
properties,
tions. While heavier leucititic lavas and travertine especially those
along with their dependence upon that influence extent and
petrographic-scale
were used in foundations and lighter pumice and chemical the
characteristics, highlights reactions
importance and, of in using
turn, determ
scoriae were sometimes used in vaults, volcanic data appropriate to either a specific tuff provenance (water-
structure and pore size distribution
tuffs (and brick fragments) comprise the majority of ratio,petrographic
or a well-documented cement chemical analogue. composition, SF
coarse aggregates in Imperial conglomeratic wall curing time and method, temperature, mix
concretes. Numerous tests by De Casa et al. (1999, etc.). In the literature various formulatio
3.3 Synopsis of Experimental
found to describe Resultsthe sorption isotherm
2007) and Jackson et al. (2005) on two commonly
used tuffs – Tufo Giallo della Via Tiberina and Tufo The accumulatedconcrete (Xi etdata
experimental al. 1994). However,
identifies sev- in th
Lionato (those found in the Great Hall concretes) – eral obstacles to paper the semi-empirical
understanding the mechanical expression
be- pro
Norling Mjornell (1997)
havior of Roman concretes. First, there is the famil- is adopted b

Proceedings of FraMCoS-7, May 23-28, 2010

iar − D (h, T )∇hwith concrete constructions in travers-
J =difficulty (1) explicitly accounts
Additionally, for the evolution
the impossibility of hydration
of obtaining suffi-
ing from the structural scale (~m) to the experimen- reactionlarge
ciently andand SFisolated
content.mortarThisspecimens
sorption from isotherm an-
tal The
scaleproportionality
(~cm). For Roman concretes,
coefficient this difficulty
D(h,T) is called reads concretes requires all tests to occur on repro-
cient
ismoisture
compounded by the fact that, given typical
permeability and it is a nonlinear function coarse ductions fabricated in the laboratory. Even when
aggregate
of the relativedimensions,
humidity ah sufficiently
and temperature representative
T (Bažant informed by the best (but⎡ inevitably limited)⎤ under-
experimental scale is on the order
& Najjar 1972). The moisture mass balance requiresof decimeters or standing of Roman materials and practices,
α s ) = G (α c α s )⎢⎢ −
we (h α cmortar 1 ⎥
+ re-
even meters. The strongly heterogeneous
that the variation in time of the water mass per unit fabric of created , ,
samples lose a (vital
1
, 1
∞ degree ⎥of accu-
g α − α c )h ⎥Further-
the concretes further complicates the
volume of concrete (water content w) be equal to the analysis of racy from differences in⎢⎣ curing e andc aging.
10
1
⎦ (4)
fracture
divergence behavior. Furthermore,
of the moisture flux J the scattered me- more, a third element in⎡ the Roman concrete
∞ − α zones com-
chanical and unknown fracture behaviors of the posite, the interfacial (g α
transition )h ⎤
(ITZs)
10
c c
coarse aggregates, combined with the length-scale between mortarKand (α α ) e
s ⎢ aggregate and −within
c coarse
⎢ ⎥
the
1
, 1

− ∂ which
= ∇ • Jthey appear in the concretes, presents(2)a ⎥
1

on pozzolanic mortar, seems⎣ to substantially influence

w
⎦
∂
majort
source of variability. Finally, the mortar is a the initiation and propagation of microcracks. While
composite
The water material
contentin wand canofbeitself. The particle
expressed as the size
sum itwhere
is notthe yetfirstknown term (gel howisotherm)
the ITZsrepresents surrounding the
distribution
of the evaporable water we (capillary water,ranges
of the Pozzolane Rosse aggregate water physically, possibly
caementa bound (adsorbed) at the same waterdecimeter
and the length-second
from
vapor,coarse silt- to medium
and adsorbed water) gravel-sized in the poz-
and the non-evaporable term (capillary
scale, affect fracture isotherm) representsthethemillimetric
propagation, capillary
zolanic
(chemically mortarsbound)
of the GreatwaterHallwnand(Mills
other monu-
1966, zones around scoriaceous mortar aggregatecontent
water. This expression is valid only for low have
ments (Jackson et al ., 2007). These
Pantazopoulo & Mills 1995). It is reasonable fragments have to of SF.observed
been The coefficient
to impactG1fracture represents the amount
trajectories (Jack- of
been
assume that the evaporable water is a functionce-
shown to redirect crack propagation in the of water
son per. 2009).
et al unit volume held in the gel pores at 100%
mentitious binding matrix
relative humidity, h, degree (Fig.of4).hydration, αc, and relative humidity, and it can be expressed (Norling
degree of silica fume reaction, αs, i.e. we=we(h,αc,αs) Mjornell 1997) as
= age-dependent sorption/desorption isotherm 4 THE FRACTURE OF IMPERIAL ROMAN
(Norling Mjonell 1997). Under this assumption and G CONCRETE c
(α α ) = k α c + k α s
c s vg c vg s
,
s (5)
by substituting Equation 1 into Equation 2 one 1

obtains The mechanical behavior, particularly in fracture, of

Roman
where kconcretes
c
vg and k vgisare
s central to understanding
material parameters. From how the
∂w ∂h
daring
maximum amount of water per unit volume thatcon-
monumental structures were initially can
− e + ∇ • ( D ∇h) = ∂we ∂w
α&c + e α&s + w&n (3) ceived, according
fill all pores (both to the empirical
capillary pores and processes
gel pores), widely one
∂h ∂t h ∂α ∂α thought
can calculateto governK1 asImperial
one obtains Roman design, and how
c s they have endured for nearly two millennia in an ac-
where ∂we/∂h is the slope of the sorption/desorption tive seismic zone on relatively⎡ poorly consolidated
⎜ g α − α ⎟h ⎥
⎛ ∞ ⎞ ⎤

isotherm (also called moisture capacity). The bedrock (Molin w − et al . 1995, Rovelli
α s + α s −G ⎢ −e
⎢
et
⎝ al. c 1995).
10
c1⎠
c s
Our experimental program ⎢⎣aims to characterize
0.188 0.22 1 ⎥

governing equation (Equation 3) must be completed ⎦ (6)

0 1
⎥

by appropriate boundary and initial conditions. the

K (α cmechanical
,α )=
s behavior on length scales that are
appropriate for the econstituent ⎜ g α − α ⎟hmaterials and ex-
⎛ ∞ ⎞
1

The relation between the amount of evaporable

10
⎝ c c⎠ −
trapolate as accurately as possible this description to
1
1

water and relative humidity is called ‘‘adsorption the The

structural scale, on whichkctheand formation of gstruc-
Figure
isotherm”4. Photomicrograph
if measured of the curving trajectory of a
debonding crack that followswith increasing
the perimeters of relativity
Pozzolane tural-scale material parameters
macrofractures that
vg ksvg andsurviving
imperil 1 can
humidity
Rosse andin ‘‘desorption
scoriae a mortar sample isotherm”
from the inGreat
the Hall,
opposite
pro- be calibrated
monuments by
occurs fitting
(Fig.experimental
5). Accordingly, data relevant
our pro-to
case. byNeglecting
duced their
a point source loaddifference (Xi et al. .1994),
test (after Jackson et al 2009). in free will
gram (evaporable)
measure water content
fracture properties inofconcrete
the diverseat
the following, ‘‘sorption isotherm” will be used with various ages (Di Luzio & Cusatis
constituents of Roman concrete and use the results 2009b).
reference to both sorption and desorption conditions. to characterize an appropriate fracture model for the
By the way, if the hysteresis of the moisture composite.
2.2 Temperature For example, evolutionthe fracture energy of the
isotherm would be taken into account, two different composite,
relation, evaporable water vs relative humidity, must G f_comp could be found to depend on the
,
Note
fracture that, at
energies earlyand age,tensile
sincestrengths
the chemical ofSFthereactions
mortar
be used according to the sign of the variation of the associated
and coarse with
aggregates,cement hydration
and the and
tensile reaction
strengths of
relativity humidity. The shape of the sorption are interfaces:
the exothermic, the temperature field is not uniform
isotherm for HPC is influenced by many parameters, for non-adiabatic systems even if the environmental
especially those that influence extent and rate of the temperature ismortar constant. mortar Heat conductionITZbrick can be
chemical reactions and, in turn, determine pore G f _ comp ξin1 Gconcrete,
described = ⋅
f + χ1 atft least for temperature
⋅ χ n ft
+ ... + ⋅ (1)
not
structure and pore size distribution (water-to-cement exceeding 100°C (Bažant & Kaplan 1996), by
ratio, cement chemical composition, SF content, Fourier’s
where ξi =law,
thewhich reads energy influence coeffi-
ith fracture
curing time and method, temperature, mix additives, cient; and χi = ith tensile strength influence coeffi-
etc.). In the literature various formulations can be cient.
q = − λThe
∇T model for the fracture properties of (7) the
found to describe the sorption isotherm of normal composite, parametrized as in equation 1 in terms of
concrete (Xi et al. 1994). However, in the present the
where component
q is the properties,
heat flux, may be T further
is the informed absolute
paper 5.the
Figure semi-empirical
Survey of macrofractures expression proposeddome
afflicting Pantheon’s by by fractographic observations
temperature, and λ is the heat conductivity; that identify additional
in this
Norling Mjornell (1997) is adopted
at the time of a restoration (after Terenzio 1934). because it physical parameters describing the fracture of the

Proceedings of FraMCoS-7, May 23-28, 2010

composite material. The model will then be opti- J = − D ( h, T of
ever, limit the applicability )∇hpoint-source results to

mized (i.e., the values for χi, ξi, determined) based understanding fracture on the structural scale. At this
on both experimental results and numerical model- scale, the formation The of fracture process zones
proportionality (FPZs) D(h,T)
coefficient
ing of the fracture processes of the cores of ancient surrounding nucleating and propagating fractures
moisture permeability and it is a nonlinea
conglomeratic concrete. Indeed, the program will could, conceivably, of thesubstantially
relative humidityreduceh and
a monu-
temperature
focus on testing numerous 20cm-diameter drill cores ment’s overall stability by weakening load
& Najjar 1972). The moisture mass paths in balanc
of the wall concrete of the Markets of Trajan, which highly stressed regions or, equivalently, reducing
that the variation in time of the water masthe
cores have been generously entrusted to our research local tensile strengths
volume of load-carrying
of concrete (waterregions.
content w) be eq
program by the Sovraintendenza Archeologica di In modern concretes,
divergence theofFPZ grows asflux
the moisture distrib-
J
Beni Culturali di Roma. As described above, the uted microcracks converge on increasingly larger
pozzolanic mortar and caementa are similar to many length scales ranging ∂ from voids in the cementitious
Imperial constructions. Ideally, the parametrized matrix (10e-6m) −to∂ the = ∇average
•J aggregate particle
w

fracture model derived from the Great Hall wall (10e-3m), depending on the local morphology and t

concretes could then be particularized to describe the stress field (van Mier The1997). As the w
water content crack
can driving
be expressed a
behavior of other Imperial concretes. energy increases,ofthe the evaporable water we in
bridging microcracks the wa
(capillary
The careful evaluation of the fracture energies of process zone eventually reach a length scale larger
vapor, and adsorbed water) and the non-e
components of the composite concrete is a first step than that of the aggregate
(chemically particles, and a water
bound) macrofrac- wn (Mil
towards the derivation of such a model. Such meas- ture may form/advance.
Pantazopoulo & Mills 1995). It is reas
urements would supplement test data already ob- The larger range of component
assume length scales
that the evaporable (up is a fu
water
tained from the concretes of the Great Hall. Jackson to 10e-1m) in Roman concretes makes the develop-
relative humidity, h, degree of hydration
et al. (2009) used point source tests on discs 3.5cm ment of the FPZdegree difficult to intuit
of silica fumequalitatively.
reaction, αs,Ini.e. we=w
in diameter and 1.5cm thick to approximate tensile stead, we introduce Hillerborg’s non-dimensional
= age-dependent sorption/desorption
strengths of the components of the wall concrete. brittleness number(Norling
to explore the process
Mjonell 1997).zone
Under quan-
this assum
The preparation of test specimens isolated specific titatively: by substituting Equation 1 into Equati
elements of the composite concrete fabric, so that the obtains
point-load tensile strengths were measured for β ( D f t 2 ) / ( E GF )
= ⋅ ⋅ (2)
caementa (brick, Tufo Lionato, Tufo Giallo della Via ∂w ∂w ∂w
Tiberina), the pozzolanic mortar, and each of the re- e ∂h + ∇dimension;
• ( D ∇h) = ft e=α& e α& + w
where D = any −structural c +
tensile
spective caementa interfaces (Fig. 6). ∂h ∂t h ∂α ∂α s
strength; E = elastic modulus; and GF = fracture c en-s
ergy. Modern dam concretes may present a rough
analog to Romanwhere concretes.
∂we/∂hTheir is thecoarse
slopeaggregate
of the sorption/
sizes approach those
isothermof typical
(also Roman moisture , capac
called caementa
while experimental measureequation
governing of their composite
(Equation 3) frac-must be
ture behavior records demonstrably
by appropriate largerandfracture
boundary initial conditi
energies compared to Thetypically-graded
relation between modern con- of e
the amount
cretes (Deng et al.water
2008).andThis agreeshumidity
relative with a general
is called ‘‘
trend of increasing fracture ifenergy
isotherm” with maximum
measured with increasing
aggregate size (Elices
humidity & Rocco 2008). In isotherm”
and ‘‘desorption modern in th
concretes, one cancase.
envision the propagating
Neglecting fracture
their difference (Xi et al.
requiring increasedtheenergy to produce
following, ‘‘sorptiona more tortu-will be
isotherm”
ous path around reference
and/or a tougher path through
to both sorption the
and desorption c
larger aggregate By particles,
the way, but ifstillthefollowing
hysteresistheof the
same general trajectory,
isothermwith would variations
be takenoccurring on two
into account,
a smaller (~mm)relation,
scale. This is not necessarily
evaporable water vs relativethe humi
case for Roman be concretes, where the
used according larger
to the signaggre-
of the varia
gate constituentsrelativity
could significantly
humidity. alter The fracture
shape of the
propagation pathsisotherm
and characteristics.
for HPC is influenced by many p
A simple two-dimensional
especially those schematic (Fig. 7)extent
that influence il- and
lustrates several possibilities.
chemical reactions The composite
and, infracture
turn, determ
Figure 6. Tensile strengths of components from the wall con- energy for cases (A-C)
structurecanandbe expressed as follows: (water-
pore size distribution
crete of the Great Hall, as measured by point source tests. ratio, cement chemical composition, SF
The point source strengths supplement, in an ap- GF A (GIc L)curing
( )
= /L time and method, temperature,
⋅ (3) mix
etc.). In the literature various formulatio
_1

proximate sense, the extremely limited data on the found to describe the sorption isotherm
tensile strength of Imperial concretes and make im- GF B (G L Gconcrete
( )
=
ITZ⋅ L (Xi
) / L et al. 1994). However,
+ ⋅ (4) in th
portant suggestions of elements in the composite
1 1 1|2 1|2

paper the semi-empirical expression pro

concrete fabric in which fractures may nucleate. The
small sample size and complicated stress fields, how- GF (G L GNorling
C
( )
=
1
⋅L GMjornell
1 ITZ
+
2
L (1997)
⋅ G L ) /isL adopted
2
+ (5)
1|3
⋅
b1|3
+
4
⋅
4

Proceedings of FraMCoS-7, May 23-28, 2010

 J = − DG
where ( hIc_1
, T )∇=h GF_1 = G1 = the fracture energy of (1) the explicitly
the ability accounts for the evolution
of the monuments to absorbofchanges
hydration
in
mortar; L1|2 = the length of the crack path along the reaction and
external andinternal
SF content.
energiesThisover sorption isotherm
many centuries.
interface between the mortar
The proportionality and a Tufo
coefficient D(h,T) Giallo della
is called reads
While much study is needed to explore this phe-
Via Tiberina caementa fragment (Fig.
moisture permeability and it is a nonlinear function8); G ITZ1|3the nomenal endurance, our initial hypothesis posits the
fracture energyhumidity
of the relative of the interface between brick
h and temperature and
T (Bažant dissipation of energy into⎡ the development of⎤ widely
mortar; and so on. In the absence of
& Najjar 1972). The moisture mass balance requires any measured distributed but relatively weakly coalesced
α s ) = G (α c α s )⎢⎢ −
we (h α c process ⎥
+ or 1

data, extremely
that the variationrough approximations
in time of the waterfor the per
mass respec-
unit bridged , ,
zones. These (gzones
1
,
∞
1
of
α − α )h ⎥ dispersed,
⎥
tive fracture energies could compute a
volume of concrete (water content w) be equal to the “composite” small-scale (with respect ⎢
⎣ eto thec meso-structure)
c ⎦ (4)
10
1

fracture
divergence energy thatmoisture
of the increasesfluxbyJaround 50% in case cracking effectively delay⎡ the localization∞ necessary
B, and almost doubles in case C, illustrating the po- for the nucleation, linkage, (g αand− αpropagation
)h ⎤
of
c c10
K (α c α s ) eperilous scales.− This
⎢ ⎥
tential of the large aggregates and interfaces to alter fractures on structurally hy-
1
, 1

− ∂ = ∇ • Jfracture properties. The estimates assume (2) ⎢ ⎥

1

composite pothesis introduces further ⎣ questions: at what ⎦ point

w

∂
a homogeneous mortar and do not take into account
t
does the accumulation of these potentially isolated
centimeter-scale
The water content scoriaw can
and belava aggregate,
expressed as thewhich
sum where the first damage
microstructural term (gelzones isotherm)
imperilrepresents the
the building
can impact fracture propagation (Figs.
of the evaporable water we (capillary water, water 4, 7d&e). physically
on bound
a structural (adsorbed)
scale? And howwater and the
suddenly? Whatsecond
role
vapor, and adsorbed water) and the non-evaporable termthe(capillary
do unusual and isotherm) represents
highly durable the capillary
alumina- and al-
(chemically bound) water wn (Mills 1966, water. This expression is valid only
kali-rich pozzolanic cements of the Roman mortars for low content
Pantazopoulo & Mills 1995). It is reasonable to of SF.in The
play coefficientofGthe
the resistance 1 represents the amount
concrete fracture? of
And,
assume that the evaporable water is a function of water globally,
more per unit volume held in the gelcomposite
could conglomeratic pores at 100%
con-
relative humidity, h, degree of hydration, αc, and relativeperhaps
cretes, humidity, and it canbybethe
characterized expressed
ability to(Norling
absorb
degree of silica fume reaction, αs, i.e. we=we(h,αc,αs) Mjornell 1997) as
energy via widely distributed, weakly coalesced
= age-dependent sorption/desorption isotherm process zones, have applications for sustainable con-
(Norling Mjonell 1997). Under this assumption and crete
G (α cconstruction
α ) = k c α cin+ seismically
s active areas? (5)
by substituting Equation 1 into Equation 2 one 1 s vg c k vg α s s
,

obtains
where kcvg and ksvg are material parameters. From the
∂w ∂h maximum amount of water per unit volume that can
− e + ∇ • ( D ∇h) = ∂we ∂w
α&c + e α&s + w&n (3) fill all pores (both capillary pores and gel pores), one
∂h ∂t h ∂α ∂α can calculate K1 as one obtains
c s
where ∂we/∂h is the slope of the sorption/desorption ⎡
⎢ 10⎜
⎛
g α c∞ − α c ⎞⎟h ⎤⎥
isotherm (also called moisture capacity). The w − 0.188 α s + 0.22α s G
c s − ⎢1 − e ⎝ 1 ⎠
⎥

governing equation (Equation 3) must be completed (6)

0 1
⎢ ⎥
K (α c α s ) = ⎣ ⎦
by appropriate boundary and initial conditions. 1
,
⎛
g αc − αc h
∞ ⎞

The relation between the amount of evaporable

10⎜ ⎟
e ⎝ 1
− ⎠ 1

water and relative humidity is called ‘‘adsorption

isotherm”
Figure if measured
7. Schematic showing with
possibleincreasing relativity
meso-scale (A-C) and The material parameters kcvg and ksvg and g1 can
humidity and
micro-scale (d-e) ‘‘desorption
mechanisms ofisotherm”
fracture in ainRoman
the opposite
concrete be calibrated by fitting experimental data relevant to
case. Neglecting their difference (Xi et al. 1994), in
vault. free (evaporable) water content in concrete at
the following, ‘‘sorption isotherm” will be used with various ages (Di Luzio & Cusatis 2009b).
It seemsto reasonable
reference both sorption to and
postulate relatively
desorption large
conditions.
fracture
By the energies
way, if for theRoman concretes.
hysteresis of theCombining
moisture
this with their
isotherm wouldlower tensile
be taken intostrengths
account,(Fig.
two 6), equa-
different 2.2 Temperature evolution
tion 2 suggests large, possibly meter-scale,
relation, evaporable water vs relative humidity, must process Note that, at early age, since the chemical reactions
zones
be usedpreceding
accordingstructural-scale
to the sign of the macrofractures
variation of the in associated with cement hydration and SF reaction
Roman
relativityconcretes.
humidity.HowThesuchshape a process
of the zonesorption
might are exothermic, the temperature field is not uniform
influence
isotherm for the HPC
propagation of fracture
is influenced by manyand parameters,
consequent for non-adiabatic systems even if the environmental
structural-scale
especially those that influence extent and rate ofisthea
destabilization of a monument temperature is constant. Heat conduction can be
complex
chemical question.
reactionsHowever,
and, in inturn,
viewdetermine
of the extraor-
pore described in concrete, at least for temperature not
dinary
structuresurvival
and pore of size
manydistribution
Imperial (water-to-cement
age monuments, exceeding 100°C (Bažant & Kaplan 1996), by
exposed to two chemical
ratio, cement millennia of differential ground
composition, sub-
SF content, Fourier’s law, which reads
sidence and seismic ground motions,
curing time and method, temperature, mix additives, we tentatively
suggest
etc.). Inherethe that the mechanical
literature strength of thecancon-
various formulations be q = − λ ∇T (7)
cretes,
found tolikely modest
describe thebut certainly
sorption sufficient,
isotherm is of
of normal Figure 8. Top: photograph of wall concrete core with compo-
secondary
concrete (Xi importance.
et al. 1994).Perhaps far more
However, relevant
in the to
present nents outlined: 1-pozzolanic mortar; 2-Tufo Giallo della Via
where 3-brick
q is fragment;
the heat4-Tufo
flux,Lionato.
T isThetheouterabsolute
the structures’ continued stability
paper the semi-empirical expression proposed by are the fracture Tiberina; core di-
energies
Norling of their concretes,
Mjornell (1997) isempirically evident init
adopted because temperature,
ameter is 20cm. and λ isschematic
Bottom: the heatof conductivity; in thisr
test design. D = 20cm;
= 5cm.

Proceedings of FraMCoS-7, May 23-28, 2010

 We begin exploring these questions with the frac- J = − D ( h,variations
behavior, incorporating T ) ∇h in component
ture testing of laboratory re-productions of Impe- properties and their relative importance in the com-
rial Roman mortar. The geometry of the Great Hall posite response. The proportionality coefficient D(h,T)
drill cores – eccentric, hollow, thin-walled cylinders moisture permeability and it is a nonlinea
– motivates a novel test design that loads arc-shaped of the relative humidity h and temperature
specimens in three-point bending (Fig. 8). The ob- ACKNOWLEDGEMENT & Najjar 1972). The moisture mass balanc
jectives are to observe the microstructures of the that the variation in time of the water mas
mortar reproductions before testing, to record the We acknowledgevolume the generous
of concreteassistance of L.w) be eq
(water content
microstructural nucleation, coalescence, and propa- Ungaro and M. divergence
Vitti from of thetheSovraintendenza
moisture flux J ai
gation of fractures on specific length scales during Beni Culturali del Comune di Roma, Ufficio Fori
testing, and to produce estimates for the fracture en- Imperiali. ∂
ergy and tensile strength of the re-fabricated mortars − w
= ∇•J
∂
based on measurements recorded on the experimen- t

tal scale. Specimens will be tested after 7, 28, 90, REFERENCES The water content w can be expressed a
180 days, and at multi-year curing periods to ob- of the evaporable water we (capillary wa
serve how the strength and fracture properties de- A.C.I. 318 Building Code Requirements for Structural
vapor, and adsorbedHills,
water) and the non-e
velop as cementitious phases advance. Details con- Concrete and Commentary. Farmington MI: ACI.
(chemically
Brune, P. & Perucchio, bound)Concrete
R. in prep. Roman water wn (Mil
Vaulting
cerning sample fabrication, testing, and data in the Great Hall Pantazopoulo & Mills
of Trajan’s Markets. Journal 1995). It is reas
of Structural
reduction, will be published along with experimental Engineering. assume that the evaporable water is a fu
results and analysis as the project proceeds. De Casa, G. & Lombardi, G. 2007.
relative Caratterih,Fisico-Meccanici
humidity, degree of hydration
del Tufo Giallo degree
della Via
of Tiberina
silica (Roma).
fume Rendiconti
reaction, αs, i.e. we=w
Lincei 18(1): 5-25.
= age-dependent sorption/desorption
De Casa, G. et al. 1999. Il Tufo Lionato Dei Monumenti
5 SOME PRELIMINARY CONCLUSIONS Romani. Geologica (Norling Mjonell
romana 35: 1-25. 1997). Under this assum
Deng, Z. et al. 2008.by substituting Equation
Comparison 1 into Equati
between Mechanical
The fracture testing of the mortar reproductions will Properties of Dam and
obtains Sieved Concretes. Journal of
contribute the first measurement of the fracture en- Materials in Civil Engineering.
ergy of Imperial Roman concrete, which is likely a Elices, M. & Rocco, C. 2008. Effect of aggregate size on the
∂w ∂hproperties of a simple
fundamental component in the formulation of a frac- fracture and mechanical e + ∇ • ( D ∇h) = ∂we αconcrete. &
∂w
e α& + w
− Mechanics
Engineering Fracture 75: 3839-3851. c +
ture description for the composite material. Past me- ∂h ∂t h ∂α
Giavarini, C. et al. 2006. Mechanical Characteristicsc of Romans ∂α s

chanical characterizations of Roman concretes have ‘Opus Caementicium.’ In S. Kourkoulis (ed.), Fracture and
focused on compressive strength. Test results show Failure of Naturalwhere
Building∂wStones. Dordrecht: Springer.
e/∂h is the slope of the sorption/
considerable scatter, which attests to the highly het- Jackson, M.D. et al. 2005. The Judicious Selection and
erogeneous nature of the composite material and the Preservation of isotherm (also called
Tuff and Travertine Buildingmoisture
Stone in capac
importance of understanding the length scales on governing equation
Ancient Rome. Archaeometry 47(3): 485-510. (Equation 3) must be
Jackson, M.D. etbyal.appropriate boundary ofand material
2009. Assessment initial conditi
which particular mechanical and fracture properties The concretes,
characteristics of ancient relation Grande
betweenAula, theMarkets
amount of e
should be measured. Still, the reported strengths, of Trajan, Rome.waterJournalandof relative
Archaeological Science
humidity is 36:
called ‘‘
however dispersed, in conjunction with the structural Lamprecht,
2481-2492.
H.O. 1984. Opus
isotherm” caementicium:
if measured Bautechnik
with der
increasing
analysis of Imperial monuments generally indicate Römer. Düsseldorf: Beton-Verlag.
humidity and ‘‘desorption
that the concretes have strength sufficient for the Molin, D. et al. 1995. Sismicità di Roma. In isotherm”
Memorie in th
static loads of the extant architectural designs ( , Descrittive della case.
Carta Neglecting their La
Geologica d'Italia: difference
Geologia (Xi di et al.
Brune & Perucchio in prep.).
c.f.
Roma I, edited by theR.following,
Funiciello, ‘‘sorption
331-407. Rome:isotherm”
Istitutowill be
Poligrafico e Zecca dello Stato.
reference to both sorption and desorption c
Mechanical analyses to determine the factors be- Rovelli, A. et al. 1995. Previsione del the
moto hysteresis
del suolo eof the
hind the survival of these designs – and the cementi- modellazione degli effetti locali. Inif Memorie
By the way, Descrittive
tious materials that preserved them while subjected isotherm
della Carta Geologica would
d'Italia: La be taken diinto
Geologia account,
Roma I, two
to centuries of seismic and subsidence events – relation, evaporable
edited by R. Funiciello,416-432. water Poligrafico
Rome: Istituto vs relative humi
requires investigation into how microcracks in the be used according to the sign of the varia
e Zecca dello Stato.
Samuelli Ferretti, A.relativity
1996. Rapporto humidity.
sulle proveThe shape of the
di laboratorio
composite concrete nucleate, propagate, and poten- ed in situ effettuate
isotherm sui componenti,
forcostruttive in
HPC is influencedMateriali bydamany p
tially resist fracture at the structural scale. The wide costruzione e especially
tecnologie those del
that influence patrimonio
extent and
variation of aggregate compositions and consequent archeologico e monumentale romano
chemical reactions con particolare
and, in turn, determ
mechanical properties of the conglomeratic compos- riferimento al tipo laziale ed all'opus latericium. Roma.
ites and their components makes the identification of structure and pore size
Terenzio, A. 1934. La restoration du Pantheon de Rome. distribution (water-
Conservation desratio,
monumentscementd’art &chemical
d’historie: composition,
280-285. SF
structural-scale material properties by direct experi- Van Mier, J.G. 1997. Fracture Processes
curing time and method, temperature,of Concrete: mix
mental testing extremely difficult. Instead, our ap- Assessment of Material
etc.). InParameters for Fracture
the literature various Models.
formulatio
proach will be to measure relevant fracture and me- Boca Raton, FL: CRC Press.
found to describe the sorption isotherm
chanical properties on the micro- and meso- scales. concrete (Xi et al. 1994). However, in th
These data will inform a parametrized model for the paper the semi-empirical expression pro
composite that will aim to create a reasonably Norling Mjornell (1997) is adopted b
bounded envelope for the structural-scale fracture
Proceedings of FraMCoS-7, May 23-28, 2010