Mechanical Behavior of Cor Tuf Ultra Hig

ERDC/GSL TR-18-31
Mechanical Behavior of Cor-Tuf Ultra-High

Performance Concrete Considering Aggregate
and Paste Effects
Isaac L. Howard, Ashley Carey, Megan Burcham, October 2018
Dylan A. Scott, Jameson D. Shannon, Robert D. Moser,
and Mark F. Horstemeyer
Geotechnical and Structures Laboratory
Approved for public release; distribution is unlimited.

The U.S. Army Engineer Research and Development Center (ERDC) solves
the nation’s toughest engineering and environmental challenges. ERDC develops
innovative solutions in civil and military engineering, geospatial sciences, water
resources, and environmental sciences for the Army, the Department of Defense,
civilian agencies, and our nation’s public good. Find out more at
www.erdc.usace.army.mil.
To search for other technical reports published by ERDC, visit the ERDC online library
at http://acwc.sdp.sirsi.net/client/default.
ERDC/GSL TR-18-31
October 2018
Mechanical Behavior of Cor-Tuf Ultra-High

Performance Concrete Considering Aggregate
and Paste Effects
Dylan Scott, Robert Moser, and Jameson D. Shannon
Geotechnical and Structures Laboratory
Concrete Materials Branch
U.S. Army Engineer Research and Development Center (ERDC)
3909 Halls Ferry Road
Vicksburg, MS 39180
Isaac L. Howard
Center for Advanced Vehicular Systems and
Civil and Environmental Engineering Department
Mississippi State University
200 Research Boulevard
Starkville, MS 39759
Ashley Carey, Megan Burcham, and Mark F. Horstemeyer

Center for Advanced Vehicular Systems and
Mechanical Engineering Department
Mississippi State University
200 Research Boulevard
Starkville, MS 39759
Final report
Prepared for U.S. Army Corps of Engineers

Washington, DC 20314-1000
Under ERS, IVPPED, and WD63 multiscale cements contracts
ERDC/GSL TR-18-31 ii
Abstract
This research primarily focused on properties from varying the
constituents that make up ultra-high performance concrete (UHPC) with
the ultimate goal to enable improved characterization and modeling of this
material. Several variations of UHPC were made to see the differences in
properties as a function of constituents. Compressive strength, elastic
modulus, and tensile strength were measured at low loading rates.
Fundamental test methods were used for most experiments with a smaller
subset of tests with strain gages and imaging techniques. This report is
intended primarily to document these experiments and the collected data.
Specific conclusions are avoided herein, as the intent is to use these data in
future efforts that will be more appropriate to draw more meaningful
conclusions about ways to better model and ultimately improve UHPC.
DISCLAIMER: The contents of this report are not to be used for advertising, publication, or promotional purposes.
Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products.
All product names and trademarks cited are the property of their respective owners. The findings of this report are not to
be construed as an official Department of the Army position unless so designated by other authorized documents.
DESTROY THIS REPORT WHEN NO LONGER NEEDED. DO NOT RETURN IT TO THE ORIGINATOR.
ERDC/GSL TR-18-31 iii
Contents
Abstract .......................................................................................................................................................... ii
Figures and Tables .........................................................................................................................................v
Preface ...........................................................................................................................................................vii
Unit Conversion Factors ........................................................................................................................... viii
1 Introduction ............................................................................................................................................ 1
1.1 Background .................................................................................................................... 1
1.2 Objectives ...................................................................................................................... 1
1.3 Scope ............................................................................................................................. 2
1.4 Symbols and acronyms ................................................................................................. 2
2 Experimental Program ......................................................................................................................... 7

2.1 Constituent material properties.................................................................................... 7
2.2 Specimen preparation and curing ................................................................................ 9
2.3 Mixtures tested ............................................................................................................ 13
2.4 Mechanical property test methods............................................................................. 15
2.4.1 Compressive strength testing (non-instrumented).................................................. 15
2.4.2 Elastic modulus testing (non-instrumented) ........................................................... 16
2.4.3 Tensile strength testing (non-instrumented) ........................................................... 17
2.4.4 Compressive and tensile testing instrumented with strain gages ......................... 18
2.5 Volume fractions .......................................................................................................... 19
2.5.1 As-batched volume estimations ............................................................................... 19
2.5.2 Volume fractions from imaging techniques ............................................................. 20
2.6 Test matrix ................................................................................................................... 23
3 Test Results ..........................................................................................................................................25

3.1 Varying proportions ..................................................................................................... 25
3.2 Curing effects .............................................................................................................. 28
3.2.1 Curing effects measured on laboratory molded specimens ................................... 28
3.2.2 Curing assessment of cores and laboratory molded specimens ........................... 36
3.3 Size effects .................................................................................................................. 37
3.3.1 Size effects – Ultra-High Performance Concrete (UHPC) ........................................ 37
3.3.2 Size effects – Cement paste (CP) ............................................................................. 38
3.3.3 Size effects – Fiber Reinforced Paste (FRP) ............................................................ 39
3.4 Instrumentation test results ....................................................................................... 42
3.5 Volume fractions estimated from batch quantities ................................................... 49
3.6 Volume fractions estimated from imaging ................................................................. 50
4 Summary ...............................................................................................................................................52
References ...................................................................................................................................................54
ERDC/GSL TR-18-31 iv
Appendix A: Strain Gage Locations .........................................................................................................56
Report Documentation Page

ERDC/GSL TR-18-31 v
Figures and Tables

Figures
Figure 1. (a) Mixing dry materials and water, (b) adding admixture, (c) finished mixture, (d)
vibration table, (e) specimens in curing room, (f) specimens in water bath, (g) end
grinding, and (h) shelved specimens awaiting testing. ...........................................................................11
Figure 2. UHPC tub provided by ERDC (a) before, (b) during, and (c) after coring. .............................14
Figure 3. Representative compressive strength test...............................................................................16
Figure 4. Representative elastic modulus testing. .................................................................................. 17
Figure 5. Representative tensile testing. ..................................................................................................18
Figure 6. (a) Compressive testing, (b) elastic modulus testing, and (c) tension testing with
strain gages...................................................................................................................................................19
Figure 7. ImageJ process for turning (a) scanning electron microscope images to a (b)
binary image and then (c) isolating inclusions......................................................................................... 21
Figure 8. Geometries generated using the area fractions and length scales found through
image analysis. .............................................................................................................................................23
Figure 9. Average compressive strength for 791 oC-d cured specimens. ............................................26
Figure 10. Average tensile strength for 791 oC-d cured specimens...................................................... 27
Figure 11. Average compressive strength of curing variations. ............................................................30
Figure 12. Average tensile strength of curing variations. ...................................................................... 31
Figure 13. 721 oC-d curing versus 791oC-d curing in compression. .................................................... 31
Figure 14. 801 oC-d curing versus 791 oC-d curing in compression. ...................................................32
Figure 15. 2,783 oC-d curing versus 791 oC-d curing in compression. ...............................................32
Figure 16. 721 oC-d curing versus 791 oC-d curing in tension. .............................................................33
Figure 17. 801 oC-d curing versus 791 oC-d curing in tension...............................................................33
Figure 18. 2,783 oC-d curing versus 791 oC-d curing in tension. .........................................................34
Figure 19. Compressive strength of CP-0.15 size variation specimens. ..............................................39
Figure 20. Compressive strength of FRP variation specimens. ............................................................ 41
Figure 21. Tensile strength of FRP variation specimens. ......................................................................42
Figure 22. Percent of ultimate compressive stress versus strain for (a) UHPC, (b) CP-0.15,
(c) M-0.56, and (d) FRP-0.23. .....................................................................................................................43
Figure 23. Percent of ultimate tensile stress versus strain for (a) UHPC, (b) CP-0.15, (c) M-
0.56, and (d) FRP-0.23................................................................................................................................45
Figure 24. Truncated stress versus vertical strain plots for (a) UHPC, (b) CP-0.15, (c) M-
0.56, and (d) FRP-0.23 in compression.................................................................................................... 47
Figure 25. Poisson’s ratio results from compression testing. ................................................................49
Figure 26. CT scan of UHPC showing (a) approximately 3% steel fibers and (b) large voids
illustrated....................................................................................................................................................... 51
Figure A1. Strain gage locations of UHPC in compression. ....................................................................56
Figure A2. Strain gage locations for CP in compression. ........................................................................ 57
Figure A3. Strain gage locations for M in compression. .........................................................................58
Figure A4. Strain gage locations for FRP in compression. ......................................................................59
ERDC/GSL TR-18-31 vi
Figure A5. Strain gage locations for UHPC in tension. ............................................................................60

Figure A6. Strain gage locations for CP in tension................................................................................... 61
Figure A7. Strain gage locations for M in tension.....................................................................................62
Figure A8. Strain gage locations for FRP in tension. ...............................................................................63
Tables
Table 1. Properties of constituent materials. ............................................................................................. 8
Table 2. Cement properties from mill certificates. .................................................................................... 9
Table 3. Batching quantities for 0.1 ft3 (2,832 cm3) of Cor-Tuf (baseline condition). .........................10
Table 4. Batch proportions and identifiers for mixtures tested. ............................................................14
Table 5. Testing matrix .................................................................................................................................24
Table 6. Varying proportions test results...................................................................................................25
Table 7. Elastic modulus and ultimate strength constant for varying proportions. .............................28
Table 8. Curing effects test results. ...........................................................................................................29
Table 9. Elastic modulus and ultimate strength constant for varying curing. .....................................36
Table 10. Size effects on CP data. ............................................................................................................39
Table 11. Size effects of FRP-0.23 in compression................................................................................40
Table 12. Size effects on FRP-0.23 in tension. .......................................................................................42
Table 13. Strain gage data for compression and tension.......................................................................46
Table 14. Comparison of elastic moduli from strain gages and compressometer. ............................48
Table 15. Average volume fractions estimated from batch quantities. ................................................49
Table 16. Comparison of average UHPC volume fractions from batching and ImageJ. .....................50
ERDC/GSL TR-18-31 vii
Preface
This study was conducted for USACE through a collaboration between the
U.S. Army Engineer Research and Development Center (ERDC) and the
Mississippi State University (MSU) Center for Advanced Vehicular
Systems (CAVS). The contents of this report were supported by three
research contracts: “Engineered Resilient Systems” (ERS), “Integrated
Virtual Prototyping for Product Engineering and Design” (IVPPED), and
“Engineering Work Directive 0063” (WD63). IVPPED was sponsored by
ERDC under Cooperative Agreement Number W56HZV-17-C-0095. The
technical monitor for IVPPED was Ms. Vernessa Noye. Dr. Roger L. King
was the CAVS Director and Principal Investigator for most activities
documented in this report, and Dr. Clay Walden assumed the
responsibility of CAVS Director near the end of this report’s activities.
The work was performed by the Concrete and Materials Branch (CMB) of
the Engineering Systems and Materials Division (ESMD), ERDC
Geotechnical and Structures Laboratory (GSL). At the time of publication,
Mr. Christopher Moore was Chief, CMB; Dr. Gordon W. McMahon was
Chief, ESMD; and Dr. Michael Sharp was Technical Director for Civil
Works Infrastructure. The Deputy Director of ERDC-GSL was
Dr. William P. Grogan, and the Director was Mr. Bartley P. Durst.
COL Ivan P. Beckman was the Commander of ERDC, and Dr. David W.
Pittman was the Director.
ERDC/GSL TR-18-31 viii
Unit Conversion Factors

Multiply By To Obtain
degrees Fahrenheit (F-32)/1.8 degrees Celsius
inches 0.0254 meters
pounds (force) per square inch 6.894757 kilopascals
pounds (mass) 0.45359237 kilograms
pounds (mass) per cubic inch 2.757990 E+04 kilograms per cubic meter
square inches 6.4516 E-04 square meters
ERDC/GSL TR-18-31 1
1 Introduction
1.1 Background
In recent years, ultra-high performance concrete (UHPC) has been studied
from many perspectives including flexural and tensile properties (Roth
2008), fiber size and shape (Scott et al. 2015), and nanomechanical
analysis of calcium-silicate-hydrate (Chandler et al. 2012). However,
studying effects of individual constituents and, for example, how they fit
into a multiscale numerical modeling framework have not been as widely
studied.
UHPC is usually composed of five to eight individual materials with

varying sizes and length scales. Typical materials are in one of five
categories: water, cementitious material, admixture, fiber, or fine
aggregate. Coarse aggregates are not incorporated into UHPC to avoid
relatively large failure planes that can occur when large aggregates have a
weak plane within them. This problem is lessened when only fine
aggregates, such as sand, are incorporated. UHPC makes use of water
reducing admixtures so less water is needed for equal fluidity. With less
water, cement paste (CP) is stronger after hydration. Fibers, typically steel,
but in some cases polymeric are added for improved ductility.
UHPC is not used nearly as often as traditional concrete, but it does have
several applications. This report does not contain a comprehensive
literature review as its primary purpose is to document a series of
fundamental experiments that are intended to be utilized in companion
efforts. A state-of-the-art report on UHPC and its history is available in
Green et al. (2014).
1.2 Objectives
The data collected and documented in this report aim to further the UHPC
knowledge base for improved characterization and are also intended for
use in numerical modeling. A laboratory test program undertaken to
measure fundamental UHPC properties, beginning with individual
constituents, is documented here. The scope of this effort focuses mainly
on experimental efforts with Cor-Tuf UHPC, which was developed by
ERDC/GSL TR-18-31 2
ERDC. An ultimate goal is for this report to be used alongside additional

data in high-performance computing environments.
1.3 Scope
This effort investigated properties at low load rates and with no specimen
confinement, which allowed more specimen replication, albeit producing
less sophisticated measurements. The goal of this portion of a larger
multiscale cementitious materials program was to understand the effects
of individual UHPC components. Once fundamental understanding of
individual components, proportioning, and their interfacial behavior is
better understood, mathematical relations to improve the state-of-the-art
in numerical modeling of UHPC should be more feasible.
Four types of specimens were evaluated: CP, mortar (M), fiber reinforced
paste (FRP), and UHPC. CP has no aggregates or fibers, M has no fibers,
FRP has no aggregates, and UHPC has all ingredients. A systematic
evaluation containing these specimen types produced from the same
ingredients (sometimes with varying proportions) is not commonplace
and is the main contribution of this effort. Specimens of these varying
types are intended to isolate contributions/behaviors of the three primary
ingredient classifications (paste, aggregates, and fibers) that can be
benchmarked to these materials combined (i.e., Cor-Tuf UHPC).
This report represents work performed beginning in 2013 under three

programs: Engineered Resilient Systems (ERS), Integrated Virtual
Prototyping for Product Engineering and Design (IVPPED), and
Engineering Work Directive 0063 (WD63). Earlier efforts within ERS
were largely conceptual, with emphasis on how laboratory measurements
of individual constituents might be able to support high performance
computing based modeling and simulation investigations focused on
concrete multi-scale modeling. Later, specimens began to be prepared,
cured, and tested (mostly within IVPPED). Thereafter, data analysis
occurred mostly within WD63. Report assembly occurred progressively
throughout the work.
1.4 Symbols and acronyms

The following symbols and acronyms are used throughout the report.
ACI – American Concrete Institute

ERDC/GSL TR-18-31 3
API – American Petroleum Institute
ASTM – American Society for Testing and Materials
C469 – elastic modulus constant found from the relationship of

compressive strength to elastic modulus
CAVS – Center for Advanced Vehicular Systems
CMRC – Construction Materials and Research Center
COV – coefficient of variation, standard deviation divided by mean
CP – cement paste
CP-0.11 – cement paste with a water to cementitious material ratio of 0.11
CT – computed topography
D – diameter
E – elastic modulus
E469 – elastic modulus from ASTM C469
ESG – elastic modulus measured by using strain gages to measure strain
ERDC – U.S. Army Engineer Research and Development Center
ERS – Engineered Resilient Systems
FRP – fiber reinforced paste
FRP-0.11 – fiber reinforced paste with fibers to cementitious material ratio

of 0.11
ERDC/GSL TR-18-31 4
FRP-0.17 – fiber reinforced paste with fibers to cementitious material ratio

of 0.17
FRP-0.23 – fiber reinforced paste with fibers to cementitious material

ratio of 0.23
HPC – High Performance Computing
HSR – High Sulfate Resistance
IVPPED – Integrated Virtual Prototyping for Product Engineering and

Design
M – mortar
M-0.47 – mortar with a fine aggregate to cementitious material ratio of 0.47
MPa – megapascal
MSU – Mississippi State University
N - Newton
oC-d – measure of maturity that is degree Celsius * days
P – load
RVE – Representative Volume Element
SEM – Scanning Electron Microscope
SG – strain gage
St – tensile strength
UHPC – Ultra-high performance concrete

ERDC/GSL TR-18-31 5
Va – volume of air
Vam – volume of admixture
Vc – volume of cement
Vnoair – volume of the specimen if there were no air
Vs – volume of sand
Vsf – volume of silica fume
Vsfl – volume of silica flour
VTotal – total volume
Vw – volume of water
WD63 – Engineering Work Directive 0063
a – air
am – admixture
c – cement
cm – cementitious material (class H cement and silica fume)
ddial – displacement of the compressometer dial
f – fibers
f/cm ratio – ratio of fibers to cementitious materials
fa – fine aggregate (silica flour and silica sand)
fa/cm – ratio of fine aggregate to cementitious materials
fc – compressive strength
ERDC/GSL TR-18-31 6
l – length
l/D – aspect ratio
mb-per – batching mass percentage of constituent as a percent
mb-tot – total mass of the batch
mcon – mass of the constituent
mcon-spec – mass of each constituent in the specimen
mspec – total mass of the specimen
mm - millimeter
psi – pounds per square inch
s – sand
sf – silica fume
sfl – silica flour
w – water
w/cm – water to cementitious materials ratio
γ – specific gravity
ε – strain
εmax – highest ultimate tensile or compressive strain
ν – Poisson’s ratio
%UltS – percent of ultimate stress

ERDC/GSL TR-18-31 7
2 Experimental Program
Cor-Tuf ultra-high performance concrete (UHPC) is composed of eight
different volume constituents that have reported compressive strengths
approaching 32,000 psi (220 MPa). An experimental program was
undertaken, as documented in this report, in which 254 specimens were
produced from Cor-Tuf constituents.
Testing, performed on the Mississippi State University (MSU) campus,

made use of raw materials provided by ERDC. Some specimens were made
and cured according to protocols used at ERDC. In addition to the ERDC
protocols, variations in curing, constituent proportions, and specimen size
were also performed.
Standardized test methods, such as those recommended by ASTM

International, are widely used and can provide some insight into
numerical modeling, but often they are not intended to provide a full
description of a material’s behavior. As such, this report deviated from
traditional test methods in some areas, as the data contained are
ultimately envisioned for use with numerical models.
Specimens were tested for several properties. Tests included compressive

strength according to ASTM C39 (ASTM International 2016b), elastic
modulus according to ASTM C469 (ASTM International 2014), and tensile
strength according to ASTM C496 (ASTM International 2011b). Some
specimens were also tested while instrumented with strain gages. Volume
fractions were estimated by volumetric principles, imaging programs
(ImageJ, Abaqus CAE 2014), and a computed topography (CT) scan.
2.1 Constituent material properties

Eight different constituents are used to make Cor-Tuf UHPC (Table 1). The
remainder of this section summarizes properties provided within
manufacturer literature relative to six of the eight constituents. Air (a) is a
volumetric constituent for which no data are necessary, and tap water (w)
was used as needed throughout the project where the specific gravity (γ)
was taken as 1.00. In this report, cementitious material (cm) was defined
as cement (c) plus silica fume (sf) on a mass basis. Also, fine aggregate (fa)
was defined as silica flour (sfl) plus sand (s).
ERDC/GSL TR-18-31 8
Table 1. Properties of constituent materials.

Constituent Abbreviation Category γ1 Description
Air a --- 0.00 Air filling voids in UHPC
Water w --- 1.00 Taken from laboratory tap
Fibers f --- 7.85 Dramix® 3D 55/30 BG
Admixture am --- 1.08 ADVA® 190
Cement c cm 3.15 API Class H (HSR) Cement
Silica Fume sf cm 2.25 Elkem Microsilica ES 900-W
Silica Flour sfl fa 2.65 SIL-CO-SIL® 75 Ground Silica
Sand s fa 2.65 F-50 Whole Grain Silica
1 γ = apparent specific gravity.
Fibers (f) that were made with low carbon or mild steel and conformed to
ASTM A820 (ASTM 2016a) were taken from Bekaert. These fibers have a
length (l) of 3 cm, a 0.55-mm diam (D), and an aspect ratio (l/D) of 55.
The fibers have a three-dimensional (3-D) geometry, a tensile strength of
approximately 1.35 N/mm2, and an elastic modulus (E) of approximately
210 N/mm2. The primary roles of fibers within UHPC are ductility and
tensile property improvements.
The admixture (am) was obtained from W.R. Grace and conformed to
ASTM C494 (ASTM 2017a) Type A and F and to ASTM C1017 Type I
(ASTM 2013). The admixture is a polycarboxlate-based high-range water
reducer. Use of a high-range water reducer allowed the concrete to have a
lower water-to-cementitious material (w/cm) ratio and higher strengths.
Cement (c) was supplied by Lafarge (a member of LafargeHolcim) out of

its Joppa, IL, facility. The product meets American Petroleum Institute
(API) Class H with high sulfate resistance (HSR). Table 2 summarizes
properties from February to April of 2016 mill certificates. This cement is
helpful for producing high ultimate strengths.
Silica fume (sf), obtained from Elkem (a Bluestar Company), had an SiO2
of greater than 85 percent, a ZrO2 of less than 10 percent, and a CaO of less
than 4 percent. Particle sizes were approximately 0.5 microns. Silica fume
lowers concrete permeability, increases corrosion resistance, and reduces
the transition zone between paste and aggregates, thus increasing bond
strength.
ERDC/GSL TR-18-31 9
Table 2. Cement properties from mill certificates.

Property February 2016 March 2016 April 2016 Average
Blaine fineness1 (m2/kg) 311 302 311 308
SiO2 (%) 22.0 22.0 22.3 22.1
Al2O3 (%) 2.7 2.7 2.6 2.7
Fe2O3 (%) 4.5 4.5 4.4 4.5
CaO (%) 64.6 64.6 64.7 64.6
MgO (%) 2.1 2.3 2.3 2.2
SO3 (%) 2.8 2.8 2.8 2.8
2Free lime (%)1 0.5 0.6 0.4 0.5
C3S (%) 63 63 62 63
C3A (%) 0 0 0 0
1Blaine Fineness measured via ASTM C204, and chemical properties measured via ASTM C114.
21: from X-ray diffraction (XRD)
Silica flour (sfl) was obtained from US Silica™ in Berkeley Springs, WV.
This material is an inert crystalline form of silica, or ground silica sand.
Approximately 99 percent of the material is finer than 75 microns, and
approximately 88 percenet of the material is finer than 45 microns (No.
325 sieve). Silica flour’s mineralogy is quartz with a pH of 7.0. Chemically,
the materials are approximately 99.5 percent SiO2. This material is largely
inert and can increase specimen density without adversely affecting
hydraulic or pozzolanic reactions.
Sand (s) was obtained from US Silica™ in Ottawa, IL. The material has
rounded particles of quartz mineralogy with a pH of 7.0. Approximately
98 percent of the particles pass a 425 micron (Number 40) sieve,
approximately 17 percent pass a 212 micron (No. 70) sieve, approximately
3 percent pass a 150 micron (No. 100) sieve, and all are retained on a
75 micron (No. 200) sieve. Silica sand is inert, a characteristic which allows
increased specimen density without affecting hydraulic or pozzolanic
reactions.
2.2 Specimen preparation and curing

The materials described in Section 2.1 were utilized to make four
categories of specimens for which Cor-Tuf was a baseline: UHPC, CP,
(3) M, and FRP. These specimen categories are detailed in Section 2.3.
Figure 1 provides photographs of key steps in the specimen preparation
ERDC/GSL TR-18-31 10
and curing process applicable to all four categories. All specimen

categories were prepared in essentially the same manner except that, for
example, specimens without fibers omitted appropriate steps. ERDC
provided baseline proportions and batching instructions for Cor-Tuf
UHPC (Table 3), which were used as a reference for all specimens
produced.
Table 3. Batching quantities for 0.1 ft3

(2,832 cm3) of Cor-Tuf (baseline condition).
Constituent Quantity
Water 1.03 lb (467 g)
Fibers 1.54 lb (699 g)
Admixture 35.3 mL
Cement 4.92 lb (2,232 g)
Silica Fume 1.92 lb (871 g)
Silica Flour 1.36 lb (617 g)
Sand 4.77 lb (2,164 g)
Dry materials (cement, silica fume, silica flour, and sand) were weighed
individually by using a digital scale then placed together into a plastic
bucket and lid. Fibers were batched separately into a bowl, and admixture
was poured into a graduated cylinder. Water was batched into two
containers, one containing 80 percent of the water needed and another
containing 20 percent of the water needed.
Once all materials were batched, the cement, silica fume, silica flour, sand,
and 80 percent of the water required were poured into a tabletop mixer
with paddle attachment set to a low speed (Figure 1a). Most specimens
were mixed in a Hobart HL200, whereas a few specimens were mixed in a
Hobart N50 5-quart mixer. Once cement, silica fume, silica flour, and
80 percent of the water were blended together, the admixture was added
with the remaining 20 percent of water, which is used to rinse the
graduated cylinder to ensure all admixture is incorporated (Figure 1b).
These materials were then mixed for 10-15 min or until the mixture was
“broken over” (reached a fluid self-consolidating consistency). The mixing
bowl edges were periodically scraped (before and after mixture breaking)
to ensure complete mixing. Once the mix reached a good fluidity, fibers
were slowly added. Once fibers were well dispersed, mixing was
completed. A total mixing time of about 20 min was typical (Figure 1c).
Figure 1. (a) Mixing dry materials and water, (b) adding admixture, (c) finished mixture, (d)
vibration table, (e) specimens in curing room, (f) specimens in water bath, (g) end grinding,
and (h) shelved specimens awaiting testing.
(a) (b) (c)
(d) (e) (f)
(g) (h)
The mixture was then placed into plastic cylinder molds (2- by 4-, 3- by 6-,
or 4-in. diam by 8-in. height) in two approximately equal lifts. A lift was
placed, tapped around the perimeter to consolidate the mixture and
remove entrapped air, and then vibrated. A vibrating table was used due to
the presence of fibers, as rodding can affect fiber alignment and
distribution. Vibration occurred for 1 to 2 min to remove the remaining air
after tapping (Figure 1d). Once produced, specimens were stored in plastic
molds with surfaces covered on a lab bench at ambient temperature for
24 to 30 hr prior to removal from molds via air pressure.
After mold removal, curing was performed with two methods: (1) a 100
percent humidity room nominally maintained at 70 to 77 °F (21° C to 25°C)
(Figure 1e) and (2) water baths nominally maintained at 80 or 90 °C
(Figure 1f). Water bath curing was used to replicate steam curing (i.e.,
placing specimens under a steam blanket and using a steam generator), as
past experience at ERDC has shown submersion in 90°C water reasonably
represents steam curing. Hydrated lime was not used during curing.
Specimens were exposed to one of the four curing protocols, listed in the
following paragraphs, prior to testing.
• Curing protocol 1 (791 °C-d): After being removed from the molds,
specimens were placed in the curing room for six days (144 hr).
Thereafter, specimens were placed in a water bath for 7 days (168 hr).
Timing began when the specimens were placed in the bath, as it took
only around 2 hr for the water to heat to the nominal temperature. The
water bath started at room temperature and was then heated to a
nominal 90°C to avoid thermal shock. At the conclusion of water bath
curing, the bath was turned off, and the specimens were allowed to cool
to ambient temperature while in the water. Once the water was at room
temperature, the specimens were shelved to dry. A total of just over
14 days is required for this protocol. This protocol is usually referred to
hereafter as 791 oC-days or 791 oC-d. A fairly approximate application
of the ASTM C 1074 (ASTM International 2011a) maturity concept was
used to calculate the oC-days for all curing protocols. The nominal
laboratory bench and 100 percent humidity curing room temperature
were taken as 23 oC, and the water bath temperature was taken as
90 oC (both approximate, but reasonable). As such, 23 oC * 7 days plus
90 oC * 7 days is 791 oC-d.
• Curing protocol 2 (721 oC-d): After being removed from molds,
specimens were placed in the curing room for 6 days (144 hr). After
6 days, the specimens were placed in a water bath for 7 days (168 hr).
around 2 hr for the water to heat to the nominal temperature. The
nominal 80°C to avoid thermal shock. After 7 days, the specimens were
cooled back to room temperature and shelved to dry. A total of just
over 14 days was required for this protocol. The amount of curing for
this protocol was approximated as 23 oC * 7 days plus 80 oC * 7 days =
721 oC-d.
• Curing protocol 3 (801 oC-d): After being removed from molds,

specimens were placed in the curing room for 6 days (144 hr). After
6 days, the specimens were placed in a water bath for 8 days (168 hr).
only around 2 hr for the water to heat to the nominal temperature. The
nominal 80°C to avoid thermal shock. After 8 days, the specimens were
cooled back to room temperature and shelved to dry. A total of just
over 15 days was required for this protocol. The amount of curing for
this protocol was approximated as 23 oC * 7 days plus 80 oC * 8 days =
801 oC-d.
• Curing protocol 4 (2,783 oC-d): After mold removal, specimens were
placed in the 100 percent humidity curing room for 120 days. After
120 days, the specimens were removed from the curing room and
shelved to dry. The amount of curing for this protocol was
approximated as 23 oC * 121 days = 2,783 oC-d.
After specimens were cured and air dried for several days, they were
transported to ERDC to have their ends ground to meet ASTM C39 (ASTM
International 2016b) standards (Figure 1g). Once ground, mass, average
diameter, and average height were measured for each specimen prior to
storage ahead of mechanical property testing (Figure 1h).
2.3 Mixtures tested

Ten mixtures were produced and tested as described in Table 4. One of
these mixtures was Cor-Tuf UHPC (mixture 1), while the remaining nine
mixtures were subcategories of CP, M, and FRP with varying proportions.
Table 4 shows the constituents and batch quantities that make up each
mixture produced. In CP, M, and FRP, the constituents and batch
quantities were selected to produce 0.1 ft3 batches while maintaining
desired w/cm, fa/cm, and/or f/cm ratios.
UHPC (mixture 1) was used largely as a control and frame of reference for
the other categories (CP, M, and FRP). Two types of UHPC specimens
were evaluated herein. The first was laboratory molded, as described in
Section 2.2, and is referred to hereafter as UHPC-Base or simply as UHPC.
The second type of UHPC specimen came from cores taken from a tub cast
at ERDC during the work of Scott et al. (2015). Figure 2 shows the process
of coring 12 specimens from this tub, nine of which were 3- by 6-in.
specimens and three of which were 4- by 8-in. specimens. The average
density of these specimens was 2.52 g/cm3 and ranged from 2.51 g/cm3 to
2.54 g/cm3. Three of the 3- by 6-in. cores were tested for compressive
strength (elastic modulus was measured in two cases), while the remaining
9 specimens were tested for tensile strength. These cores were not cured
beyond what they had already experienced upon arrival at MSU and what
would occur in laboratory temperature and humidity conditions. The exact
curing history of the tub was unknown.
Table 4. Batch proportions and identifiers for mixtures tested.

Proportion
ning ratios Batching quantities for 0.1 ft3 (2,832 cm3)
Mixture Silica Silica
Water Fibers Admixture Cement fume flour Sand
Identifier w/cm fa/cm f/cm (g) (g) (ml) (g) (g) (g) (g)
UHPC-
1 0.15 0.71 0.23 467 699 35 2,232 871 617 2,164
base
UHPC-
1 0.15 0.71 0.23 --- --- --- --- --- --- ---
core
2 CP-0.11 0.11 0.00 0.00 701 --- 71 4,463 1,742 --- ---
3 CP-0.15 0.15 0.00 0.00 934 --- 71 4,463 1,742 --- ---
4 CP-0.26 0.26 0.00 0.00 1,168 --- 71 4,463 1,742 --- ---
5 M-0.47 0.15 0.47 0.00 619 --- 47 2,957 1,154 817 1,125
6 M-0.56 0.15 0.56 0.00 619 --- 47 2,957 1,154 817 1,500
7 M-0.65 0.15 0.65 0.00 619 --- 47 2,957 1,154 817 1,875
8 FRP-0.11 0.15 0.00 0.11 771 577 58 3,682 1,437 --- ---
9 FRP-0.17 0.15 0.00 0.17 771 864 58 3,682 1,437 --- ---
10 FRP-0.23 0.15 0.00 0.23 771 1,153 58 3,682 1,437 --- ---
-Water (w) was the batched amount.
--Note: all specimens were lab molded except for UHPC-Core.
--Note: batching data were not documented for UHPC-Core, but were assumed comparable to UHPC-base.
Figure 2. UHPC tub provided by ERDC (a) before, (b) during, and (c) after coring.
(a) (b) (c)
The rationale behind mixtures 2 to 10 was to vary constituent materials

among categories (CP, M, and FRP) and then to vary relative proportions
of constituent materials within categories (e.g., vary w/cm for CP or f/cm
for FRP). Having a suite of test results where constituents and relative
proportions were varied with subsequent exposure to varied curing
conditions provides a pool of mechanical properties that can be used for
multi-scale modeling purposes in subsequent efforts. For CP, the water-to-
cementitious material (w/cm) ratio bracketed UHPC. For M, the fine
aggregate to cementitious material (fa/cm) ratio was progressively lowered
below that of UHPC. For FRP, the fiber-to-cementitious material (f/cm)
ratio began at the UHPC level and was progressively lowered.
2.4 Mechanical property test methods

Mechanical property testing measured compressive strength (fc), tensile
strength (St), elastic modulus (E), Poisson’s ratio (υ), and load (P) versus
strain (ε) relationships for instrumented specimens. Mechanical property
testing was performed with a Forney compression machine having a
600-kip capacity at the Construction Materials Research Center (CMRC)
at Mississippi State University (MSU).
2.4.1 Compressive strength testing (non-instrumented)
Compressive strength was measured according to ASTM C39 (ASTM

International 2016b) as shown in Figure 3. Specimens with ground ends
were placed directly into the compression machine absent pad caps, (i.e.,
the concrete was directly in contact with metal on both surfaces). The load
at failure was divided by the original cross-sectional area to determine the
compressive strength (fc). Testing time ranged from approximately 10 to
30 min, depending on specimen strength. ASTM C39’s allowable load rate
is 28 to 42 psi/sec, which for the tested 3-in.-diam specimens equates to a
load rate of 12 to 17.5 kips/min. As shown in Figure 3, this load rate was
successfully verified.
Figure 3. Representative compressive strength test.
35,000
30,000
37 psi per Second
25,000 R² = 0.99
20,000
15,000
10,000
5,000
0
100 200
Time (Seconds)
300 400 500 600 700 800 900
2.4.2 Elastic modulus testing (non-instrumented)
Elastic modulus was found for specimens with ends ground according to
ASTM C469 (ASTM International 2014). Specimens were placed into the
compression machine fitted with a compressometer (Figure 4). The
compressometer recorded specimen elastic deformation with respect to
applied load. By using Equation 1, the recorded deformation was able to be
converted to strain.
(𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 )(0.0001)
𝜀𝜀 = 8
(1)
where
ε = strain experienced by the specimen
ddial = displacement as taken from the dial during testing
The load was then divided by the original cross-sectional area of the
specimen to obtain stress. Elastic modulus was taken as the slope of the
linear portion of the stress-strain curve and is denoted E469 herein to
associate this measurement with ASTM C469 (ASTM International 2014)
and differentiate this value from elastic modulus (E) measured from other
methods. Elastic modulus testing via C469 took about 45 min per
specimen to complete.
Figure 4. Representative elastic modulus testing.
2.4.3 Tensile strength testing (non-instrumented)
Tensile strength was measured on specimens according to ASTM C496

(ASTM International 2011b; Figure 5). The specimens were placed on their
sides and had two bearing strips (1/8-in.-thick plywood) placed between
the machine and the specimen (i.e., there was always a barrier of wood
between the specimen and the metal surfaces on the top and bottom).
These thin pieces of wood helped distribute the applied load uniformly
throughout the specimen. The maximum applied load was recorded and
used to find tensile strength by using Equation 2. Since the load rate for
this test was 3.3 psi/sec or less, experiments ranged from approximately
30 to 120 min, depending on the specimen.
2𝑃𝑃
St = 𝜋𝜋𝜋𝜋𝜋𝜋
(2)
where
St = splitting tensile strength (MPa)
P = maximum applied load (N)
l = length of the specimen (mm)
D = diameter of the specimen (mm)
Figure 5. Representative tensile testing.
2.4.4 Compressive and tensile testing instrumented with strain gages
Foil strain gages (SG) are a versatile method for measuring responses that
can be used for numerical modeling purposes. Example applications range
from strain measurement on materials much less stiff than concrete, such
as geogrids and geotextiles (Warren et al. 2010), to materials besides
concrete with stiffnesses on the order of concrete, such as fiber reinforced
polymers (Howard and GangaRao 2009), to materials stiffer than
concrete, such as metal (Rushing et al. 2016).
Vishay Micro-Measurements 350 Ω general purpose gages (C2A-06-125-

Lw-350 and C2A-06-250LW-350) were attached to specimens using M-
Bond 200 glue and catalyst following instructions given by Vishay. The
strain gages were then connected to a National Instruments NI CompaqDaq
9172 chassis and NI 9237 I/O modules that recorded all of the data collected
from the gages at a rate of slightly less than 5 Hz by using a program written
in LabVIEW. Instrumented specimens were tested in compression, and
additional instrumented specimens were tested in tension (Figure 6).
Appendix A provides drawings of strain gage locations, where strain gages
are denoted SG-0 through SG-3. Elastic modulus was also found locally by
using strain gages and is reported in Section 3.4 as ESG-0 through ESG-3.
Figure 6. (a) Compressive testing, (b) elastic modulus testing, and (c) tension testing with
strain gages.
(a) (b) (c)
2.5 Volume fractions

2.5.1 As-batched volume estimations
Volume fractions were estimated by way of batch quantities and

constituent material specific gravity values. To find these volumetric
estimations, two critical assumptions were necessary. The first assumption
was that the aggregates used do not absorb water. This is reasonable
considering the fine aggregates used have less than 0.25 percent water
absorption, based on past tests conducted by ERDC on these materials.
The second assumption was that the amount of water batched remained
constant until the specimen was tested. It is important to note that the
volume fractions calculated from as-produced densities and as-batched
quantities are estimates and do not account for volume change during
hydration, nor do they provide any information on distribution and size of
air voids. All of the batching masses for each constituent were recorded as
well as the final weight of each specimen. From the batching masses, the
mass percentage of the constituents was found for each specimen by using
Equation 3.
𝑚𝑚𝑐𝑐𝑐𝑐𝑐𝑐
𝑚𝑚𝑏𝑏−𝑝𝑝𝑝𝑝𝑝𝑝 = (3)
𝑚𝑚𝑏𝑏−𝑡𝑡𝑐𝑐𝑡𝑡
where
mcon = mass of the constituent in grams
mb-tot = total mass of the batch in grams
mb-per = batching mass percentage of constituent as a percent
This percentage was multiplied by the weight of each specimen to get the
ideal weight of each constituent in each specimen shown in Equation 4.
𝑚𝑚𝑐𝑐𝑐𝑐𝑐𝑐−𝑠𝑠𝑝𝑝𝑝𝑝𝑐𝑐 = (𝑚𝑚𝑏𝑏−𝑝𝑝𝑝𝑝𝑝𝑝 )(𝑚𝑚𝑠𝑠𝑝𝑝𝑝𝑝𝑐𝑐 ) (4)
where
mb-per = batching mass percentage of constituent as a percent
mspec = total mass of the specimen in grams
mcon-spec = mass of each constituent in the specimen in grams
These masses were then divided by their specific gravity and then summed
to find a specimen volume with no air, shown in Equation 5.
𝑚𝑚𝑐𝑐𝑐𝑐𝑐𝑐−𝑠𝑠𝑠𝑠𝑠𝑠𝑐𝑐
𝑉𝑉𝑐𝑐𝑐𝑐𝑛𝑛𝑛𝑛𝑝𝑝 = ∑ 𝛾𝛾
(5)
where
mcon-spec = mass of each constituent in the specimen in grams
γ = specific gravity of the constituent
Vnoair = volume of the specimen if there was no air in g/cm3
The volume of air (Va) was then found by finding the difference between
the laboratory recorded volume of the specimen (found by taking average
heights, diameters, and masses) and the ideal volume of the specimen with
no air (Vnoair). Vnoair is the sum of cement volume (Vc), silica fume volume
(Vsf), silica flour volume (Vsfl), sand volume (Vs), water volume (Vw), fibers
volume (Vf), and admixture volume (Vam). Each of these volumes was
determined with its corresponding mcon-spec and specific gravity.
2.5.2 Volume fractions from imaging techniques
Some specimens were used for an imaging analysis to estimate volume

fractions. These volume fractions were found by taking pictures of the
specimens then using the programs ImageJ and Abaqus to find volume
fractions. Lower magnification images were taken using a ZEISS Axiovert
200 optical microscope at the Center for Advanced Vehicular Systems
(CAVS) at Mississippi State University. A scanning electron microscope

model SUPRA 40 FEG-SEM, also at CAVS, was used to take higher
magnification images. The surfaces of the specimens were ground smooth
before imaging to ensure quality pictures. The combination of the low and
high magnification microscopes allowed for images to be taken at varying
length scales to capture the size distributions of each constituent.
The image processing tool, ImageJ, was used for analyzing images
(Schneider et al. 2012). This software was utilized to identify, isolate, and
measure constituents (Figure 7). These results allowed for the
determination of the average size, number density, area fraction, and
nearest neighbor distance of each constituent. ImageJ was used on
multiple images at varying length scales to ensure accurate results were
found for each of the constituents.
Figure 7. ImageJ process for turning (a) scanning electron microscope images to a (b) binary
image and then (c) isolating inclusions
(a) (b)
500 μm
(c)
The finite element software Abaqus was used to generate representative

volume element (RVE) cubes to portray UHPC (Abaqus 2014). It was
determined that three length scales would aid modeling of UHPC
mesoscale characteristics, as there were such differences in the size of
constituents. The largest cube had sides of 35 mm, and it included only
steel fibers within the matrix. The middle length scale included large voids
and silica sand grains and had sides of 0.54 mm. The smallest length scale
was a cube with sides of 0.04 mm, and it included small voids and
unhydrated cement grains.
To best portray the material, inclusions were randomly distributed

throughout the matrix according to their measured distributions. Results
from ImageJ gave an area fraction, while Abaqus required a volume
fraction for geometry generation. The average size of each inclusion type
was assumed to be the area of a circle, and the average radius was found.
For the geometry generation, the area fraction was assumed to be the same
as the volume fraction, and the average radius size was used to create
spheres and cylinders.
A python script was written to generate these RVE cubes. This script
required inputs for average inclusion size (found using ImageJ), and
inclusions of that size were inserted into the matrix until the desired
volume fraction was reached. Plots of constituent size versus number of
occurrences revealed that using only the average constituent size left out
the range of inclusion sizes found in UHPC. The size versus number of
occurrences graph was fitted with a distribution curve, and a second script
was created that required the mean and variance of this fitted curve. The
inclusions inserted into the matrix by this second code were of sizes that
varied according to the distribution curve.
Both random generation codes required the size of the cube, as well as the
number, shape, and volume fraction of each constituent. Two shapes,
spheres and cylinders, were used in modeling the inclusions in subsequent
simulations. Spheres were used to model the voids, unhydrated cement
grains, and sand grains; cylinders were used to model the steel fibers. The
algorithms then randomly inserted and distributed these shapes, making
sure none overlapped. Figure 8 represents the RVE generations.
Figure 8. Geometries generated using the area fractions and length scales found through
image analysis.
The Phoenix X-Ray Computed Topography (CT) system with dual focus,
reaching one-micron resolution, located at CAVS, was used for 3-D imaging.
CT scans were conducted on a UHPC cube with sides of approximately
50 mm. The CT scan distinguished the larger constituents, large voids, and
steel fibers, helping to determine the 3-D characteristics of UHPC.
2.6 Test matrix

Table 5 is the test matrix for the 254 specimens produced and tested for this
report. This matrix evaluated specimens of the four consistent categories
(UHPC, CP, M, and FRP) at different specimen sizes (2 by 4 to 4 by 8) and
after different amounts of curing (721 to 2,783 oC-d) for mechanical
properties and/or volume fractions. Eight of the mechanical property
specimens were instrumented with strain gages, and several of the
compression specimens were fitted with a compressometer for elastic
modulus determination. Volume fractions were measured by imaging on the
five specimens shown in Table 5 and were also estimated on the mechanical
property specimens by way of mass proportions and specific gravities.
Table 5. Testing matrix

Mechanical Volume
property specimens fractions
Mix Curing Size Compression Tension Imaging
UHPC 2,783 oC-d 3x6 6 6 0
CP-0.151 2,783 oC-d 3x6 15 3 0
CP-0.151 2,783 oC-d 2x4 9 0 0
M-0.56 2,783 oC-d 3x6 6 3 0
FRP-0.232 2,783 oC-d 3x6 10 11 0
FRP-0.232 2,783 oC-d 4x8 2 3 0
UHPC 791 oC-d 3x6 6 3 2
CP-0.15 791 oC-d 3x6 6 3 1
CP-0.11 791 oC-d 3x6 6 3 0
CP-0.26 791 oC-d 3x6 6 3 0
M-0.56 791 oC-d 3x6 6 3 1
M-0.47 791 oC-d 3x6 6 3 0
M-0.65 791 oC-d 3x6 6 3 0
FRP-0.23 791 oC-d 3x6 6 3 1
FRP-0.17 791 oC-d 3x6 6 3 0
FRP-0.11 791 oC-d 3x6 6 3 0
UHPC 721 oC-d 3x6 6 3 0
CP-0.15 721 oC-d 3x6 6 3 0
M-0.56 721 oC-d 3x6 6 3 0
FRP-0.23 721 oC-d 3x6 6 3 0
UHPC3 801 oC-d 3x6 6 3 0
CP-0.15 801 oC-d 3x6 6 3 0
M-0.56 801 oC-d 3x6 6 3 0
FRP-0.23 801 oC-d 3x6 6 3 0
UHPC-Core Core 3x6 3 6 0
UHPC-Core Core 4x8 0 3 0
All Specimens 159 89 5
1:2x4 and 3x6 specimens were sometimes produced in the same batch for paired
comparisons.
2:3x6 and 4x8 specimens were sometimes produced in the same batch for paired
comparisons.
3: One additional specimen (shown in Figure 3) was produced for protocol
development, but was not included in the data reported in Chapter 3 (i.e., there were
7 compression specimens, but only 6 were included in Chapter 3).
3 Test Results
The Table 5 testing matrix was largely divided into behaviors of interest
for reporting purposes. Some specimen types had multiple property
measurements, and as a result, reporting the data in this manner seemed
more logical.
3.1 Varying proportions

Of the 254 specimens produced, 90 were tested after 791 oC-d curing to
assess the effects of varying proportions on fundamental properties;
10 different types of specimens were tested (Table 6). The densities of
these specimens ranged from 2.14 g/cm3 to 2.61 g/cm3, and all were 3-in.
diam by 6-in. height. Figures 9 and 10 show side-by-side comparisons of
the variations data for compressive and tensile testing.
Table 6. Varying proportions test results.
fc Range fc St St St Density Density Density

fc Avg.
Mix ID COV Range COV Avg. Range COV Avg.
(MPa) (MPa)
(%) (MPa) (%) (MPa) (g/cm3) (%) (g/cm3)
UHPC 152.4 – 12.0 175 19.5 – 14.5 23 2.467 – 1.6 2.52
203.1 26.1 2.608
CP-0.11 68.1 – 141.0 26.4 117 8.1 – 10.4 13.6 10 2.224 – 0.9 2.27
2.286
CP-0.15 77.7 – 167.7 24.6 117 9.7 – 12.8 14.3 12 2.230 – 0.7 2.25
2.275
CP-0.26 81.5 – 137.8 19.0 113 3.7 – 8.9 40.1 7 2.140 – 0.7 2.17
2.183
M-0.47 139.8 – 7.6 161 7.7 – 13.7 29.8 12 2.247 – 0.8 2.26
175.4 2.297
M-0.56 122.2 – 9.1 141 9.3 – 12.9 17.3 12 2.233 – 1.1 2.27
155.1 2.314
M-0.65 146.0 – 11.2 164 9.4 – 15.3 27.0 12 2.254 – 0.6 2.27
192.5 2.292
FRP-0.11 93.6 – 132.0 12.5 116 19.1 – 25.1 13.7 22 2.347 – 1.0 2.39
2.412
FRP-0.17 109.3 – 9.3 125 13.9 – 30.2 21 2.436 – 1.1 2.47
139.9 26.0 2.515
FRP-0.23 120.7 – 6.4 135 26.3 – 0.5 26 2.465 – 1.6 2.51
141.9 26.6 2.584
--Note: Averages include 6 fc, 3 St, and 9 densities.
UHPC had the highest compressive strength, followed by M, FRP, and CP.
CP was not sensitive to w/cm ratio changes within the range considered, M
did not respond in a progressive manner as the fa/cm ratio was changed,
and FRP increased compressive strength as the f/cm ratio increased.
Several of the results from varying proportions were not intuitive.
FRP with the highest fiber loading produced the highest tensile strength.
Overall, UHPC and FRP had comparable tensile strengths. M tensile
strength was insensitive to fa/cm ratio changes. CP behaved somewhat
erratically in tension and produced the lowest overall strength. Overall,
tensile strength behaviors were fairly intuitive.
Figure 9. Average compressive strength for 791 oC-d cured specimens.
200
Average Compressive Strength (MPa)
180
160
140
120
100
80
60
40
20
0
Figure 10. Average tensile strength for 791 oC-d cured specimens.
30
Average Tensile Strength (MPa)
25
20
15
10
Elastic modulus was also found for specimens that underwent compressive
testing in accordance to ASTM C469 (ASTM International 2014). The
relationship between elastic modulus and compressive strength is often
reported in the form of Equation 6. The constant relating elastic modulus
and compressive strength is the parameter of primary interest, and Table 7
reports all C469 findings.
𝐸𝐸469 = 𝐶𝐶469 �𝑓𝑓𝑐𝑐 (6)
where
E469 = elastic modulus from ASTM C469 (MPa; ASTM International
2014))
C469 = constant
fc = compressive strength (MPa)
Table 7. Elastic modulus and ultimate strength constant for varying proportions.
Curing Avg. E469 COV C469 COV C469
Mix ID E469 Range (MPa) (%)
(MPa) Range (%) Avg.
UHPC 791 oC-d 49,509 – 49,617 49,600 0.1 3,848 – 4,013 3.0 3,931
CP-0.15 791 oC-d 33,826 – 36,682 35,300 5.7 3,235 – 3,433 4.2 3,334
M-0.56 791 oC-d 41,547 – 45,296 43,400 6.1 3,489 – 3,637 2.9 3,563
FRP-0.23 791 oC-d 31,460 – 32,374 31,900 2.0 2,776 – 2,947 4.2 2,862
CP-0.11 791 oC-d 40,884 – 41,786 41,300 1.5 3,443 – 3,654 4.2 3,548
CP-0.26 791 oC-d 27,887 – 28,355 28,100 1.2 2,576 – 3,142 14.0 2,859
M-0.47 791 oC-d 43,830– 45,564 44,700 2.7 3,371 – 3,637 5.4 3,504
M-0.65 791 oC-d 45,104 – 45,764 45,434 1.0 3,564 – 3,726 3.1 3,645
FRP-0.17 791 oC-d 32,116 – 38,420 35,300 12.6 2,873 – 3,674 17.3 3,273
FRP-0.11 791 oC-d 30,278 – 34,821 32,500 9.9 2,811 – 3,600 17.4 3,206
--Note: Above averages include 2 E469 values and 2 C469 values.
The American Concrete Institute (ACI) document 318 allows the constant
relating elastic modulus to compressive strength in the form of Equation 6
to be 57,000 for U.S. customary units (i.e., psi). For SI units (i.e., MPa),
this constant is approximately 4,700. The main observation from Table 7
is that the materials evaluated do not produce as high of an elastic
modulus per unit of compressive strength as typical ready-mixed concrete.
3.2 Curing effects

Of the 254 specimens produced, 153 were utilized to assess curing effects
on mechanical properties. Most of the data were from laboratory molded
specimens, with a small assessment from nine cores. Laboratory molded
data were reported first, followed by a brief assessment with the core test
results.
3.2.1 Curing effects measured on laboratory molded specimens
Of the 254 specimens produced, 144 were utilized to assess curing effects
on mechanical properties from laboratory molded specimens (Table 8).
Note that 791 oC-d data are replicated between Table 6 and Table 8.
Table 8 utilized six compression and three tensile specimens per mix at a
given type of curing. All nine of these specimens were used to assess
density, which ranged from 2.19 to 2.64 g/cm3.
Table 8. Curing effects test results.

fc fc fc St St St Density Density Density
Mix ID Curing Range COV Avg. Range COV Avg. Range COV Avg.
(MPa) (%) (MPa) (MPa) (%) (MPa) (g/cm3) (%) (g/cm3)
UHPC 152.4 – 12.0 175 19.5 – 14.5 23 2.467 – 1.6 2.52
203.1 26.1 2.608
CP-0.15 77.7 – 167.7 24.6 117 9.7 – 12.8 14.3 12 2.230 – 0.7 2.25
791 2.275
oC-d
M-0.56 122.2 – 9.1 141 9.3 – 12.9 17.3 12 2.233 – 1.1 2.27
155.1 2.314
FRP- 120.7 – 6.4 135 26.3 – 0.5 26 2.465 – 1.6 2.51
0.23 141.9 26.6 2.584
UHPC 156.2 – 7.1 170 22.1 – 15.5 24 2.487 – 0.8 2.53
183.6 28.8 2.554
CP-0.15 106.4 – 18.7 124 4.5 – 11.1 41.6 9 2.223 – 0.6 2.24
721 170.1 2.261
oC-d
M-0.56 138.6 – 15.4 179 11.0 – 7.2 12 2.262 – 1.4 2.30
208.4 12.6 2.347
FRP- 88.3 – 120.5 12.9 105 17.3 – 17.3 20 2.460 – 1.1 2.52
0.23 23.6 2.547
UHPC 155.5 – 14.5 190 18.7 – 12.7 21 2.453 – 1.8 2.51
221.5 23.8 2.605
CP-0.15 108.9 – 20.9 135 4.0 – 10.5 44.6 8 2.191 – 1.1 2.23
801 167.5 2.265
oC-d
M-0.56 120.4 – 18.3 161 10.8 – 13.0 13 2.255 – 0.9 2.28
187.4 14.0 2.309
FRP- 74.9 – 145.9 20.4 119 22.2 – 6.0 24 2.479 – 2.1 2.56
0.23 24.9 2.644
UHPC 131.5 – 7.5 149 18.8 – 8.9 21 2.527 – 0.7 2.55
163.4 22.3 2.577
CP-0.15 78.0 – 143.4 21.0 109 4.7 – 10.1 47.6 7 2.188 – 1.6 2.24
2,783 2.274
oC-d
M-0.56 126.3 – 6.1 136 10.3 – 8.2 11 2.253 – 0.9 2.28
144.8 12.0 2.310
FRP- 113.9 – 6.6 122 19.6 – 17.0 24 2.457 – 2.0 2.53
0.23 135.8 27.7 2.623
--Note: Above averages include six fc, three St, and nine densities.
As seen in Table 5, there were often more than six compression and three
tension specimens tested after 2,783 oC-d curing. In these cases, the first six
compression measurements and/or the first three tension measurements
were utilized in Table 8 for consistency across all curing protocols. The
additional replicates for 2,783 oC-d curing were performed for investiga-
tions reported later in this chapter. Figures 11 to 18 plot the results from
Table 8. Figures 11 and 12 are bar charts comparing curing effects on tensile
and compressive strength. Figures 13 to 18 are equality plots comparing
curing protocols 721, 801, and 2,783 oC-d to ERDC’s baseline protocol of
791 oC-d, where the specimens are exposed to 90 oC water.
Figure 11. Average compressive strength of curing variations.

Figure 12. Average tensile strength of curing variations.
Figure 13. 721 oC-d curing versus 791oC-d curing in compression.

Figure 14. 801 oC-d curing versus 791 oC-d curing in compression.
Figure 15. 2,783 oC-d curing versus 791 oC-d curing in compression.
Figure 16. 721 oC-d curing versus 791 oC-d curing in tension.
Figure 17. 801 oC-d curing versus 791 oC-d curing in tension.
Figure 18. 2,783 oC-d curing versus 791 oC-d curing in tension.
Curing protocols noticeably impacted compressive strength for all

specimen types. Curing protocols noticeably affected UHPC, CP, and FRP
tensile strength, but did not noticeably affect M.
Figures 13 and 14 were used to make the following assessments regarding

compressive strength. UHPC seemed to respond better to 80 oC curing
than 90 oC curing for a comparable level of oC-days. This same trend was
observed with M and CP, but not with FRP. FRP clearly deviated from the
other categories for 80 oC curing relative to 90 oC curing, as 90 oC curing
produced higher strengths.
Figure 15 was used to make the following assessments about compressive

strength. In compression, all UHPC and FRP specimens were stronger
after 791 oC-d of curing, for which 90 oC was utilized, relative to 3.5 times
the amount of oC-days (2,783), for which temperature was 23 oC. CP was
comparable, where M might have slightly favored 90 oC curing for less oC-
days. Overall, the data were below the equality line, indicating curing at
higher temperatures for less oC-days produced higher compressive
strength than lower temperatures for more oC-days.
Figures 16 and 17 were used to make the following assessments about

tensile strength. UHPC’s response was not definitive. The 721 oC-d
specimens with 80 oC curing were modestly stronger than the 791 oC-day
specimens with 90 oC curing, suggesting 80 oC curing was more favorable.
The opposite behavior, however, was observed when 801 oC-d at 80 oC
curing was compared to 791 oC-d at 90 oC curing. CP favored 90 oC curing,
which was opposite to compressive strength testing in Figures 13 and 14.
M seemed largely indifferent to curing protocol when tested in tension.
FRP favored 90 oC curing, which agrees with the behavior observed in
compression (Figures 13 and 14).
Figure 18 was used to make the following assessments about tensile

strength. UHPC favored 90 oC curing, which matched its behavior in
compression in Figure 15 (i.e., that fewer oC-days of a higher temperature
produced better properties). CP also favored 90 oC curing, which disagreed
with its compressive strength behavior. M seemed largely indifferent to the
curing protocol. FRP generally favored 90 oC curing.
The main observation from Table 8 (and Figures 11 to 18 that plotted data
from Table 8) is that the specific aspects of curing seem to affect
compressive and tensile properties and that the effects are not consistent
between tension and compression or between specimen category (UHPC,
CP, M, or FRP). These data suggest that only accumulating oC-days and
using that number as a maturity index may lead to undesirable outcomes
for some types of UHPC endeavors. For large placements of UHPC where
very large temperature fluctuations with space and time are expected,
more understanding of how UHPC responds to time, temperature, and
their interaction seems useful. This same understanding is needed to
further develop multiscale models of high-strength concrete.
Curing effects on elastic modulus are shown in Table 9. The main

observation is that the materials evaluated do not produce as high of an
elastic modulus per unit of compressive strength as typical ready-mixed
concrete.
Table 9. Elastic modulus and ultimate strength constant for varying curing.
E469 E469 E469
C469 C469
Mix ID Curing Range COV Avg. COV (%)
Range Avg.
(MPa) (%) (MPa)
UHPC 49,509 – 49,617 0.1 49,600 3,848 – 4,013 3.0 3,931
CP-0.15 791 33,826 – 36,682 5.7 35,300 3,235 – 3,433 4.2 3,334
oC-d
M-0.56 41,547 – 45,296 6.1 43,400 3,489 – 3,637 2.9 3,563
FRP-0.23 31,460 – 32,374 2.0 31,900 2,776 – 2,947 4.2 2,862
UHPC 51,810 – 57,633 5.3 54,800 4,059 – 4,611 7.5 4,246
CP-0.15 721 31,699 – 37,223 8.7 35,200 3,073 – 3,425 5.8 3,293
oC-d
M-0.56 44,626 – 50,060 5.7 47,500 3,467 – 3,505 0.6 3,484
FRP-0.23 29,684 – 31,196 2.5 30,400 2,890 – 3,206 5.5 3,085
UHPC 51,633 – 54,474 2.7 52,900 3,660 – 3,747 1.2 3,706
CP-0.15 801 35,747 – 39,371 4.9 37,400 3,034 – 3,390 6.2 3,163
oC-d
M-0.56 44,095 – 47,617 4.1 45,500 3,221 – 3,503 4.8 3,409
FRP-0.23 28,268 – 31,828 5.9 30,100 2,579 – 2,924 6.3 2,757
UHPC 48,269 – 50,001 1.9 49,000 3,922 – 4,209 3.7 4,044
CP-0.15 2,783 31,284 – 31,715 0.7 31,500 2,638 – 3,542 15.4 3,024
oC-d
M-0.56 42,339 – 43,389 1.4 42,700 3,524 – 3,718 2.7 3,616
FRP-0.23 32,299 – 37,793 8.4 34,500 2,959 – 3,513 14.2 2,918
--Note: 791 oC-d averages for E469 and C469 include 2 specimens each. All other curing averages include 3 E469
and 3 C469.
3.2.2 Curing assessment of cores and laboratory molded specimens
The 3- by 6-in. UHPC cores that were tested for compressive strength were
used as a general curing reference relative to the four laboratory protocols
utilized herein. These three cores produced an average compressive
strength of 157 MPa and an average E469 of 55,900 MPa. These values were
benchmarked with UHPC laboratory cast material into 3- by 6-in. cylinders
that were reported earlier in this section. The range of average compressive
strength and E469 were 149 to 190 MPa and 48,300 to 57,600, respectively.
The 3- by 6-in. UHPC cores that were tested for tensile strength were used
as a general curing reference relative to the four laboratory protocols
utilized herein. These six cores produced an average tensile strength of
19.8 MPa. These values were benchmarked with UHPC laboratory cast
material into 3- by 6-in. cylinders that were reported earlier in this section.
The range of average tensile strength was 20.8 to 24.4 MPa.
Variability was compared for the two available sets of UHPC tensile
strength data cured in different manners where six replicates were
available for 3- by 6-in. specimens. The 2,783 oC-d cured specimens had a
COV of 20.6 percent. Cores had a COV of 17.3 percent.
3.3 Size effects

Of the 254 specimens produced, 69 were used to assess specimen size
effects on measured properties of UHPC, CP-0.15, and FRP-0.23. The data
contained in this section are sporadic. The best available information was
collected herein with materials and resources remaining after proportions
and curing effects were assessed. The effects of size on UHPC were found
by comparing cores of two different diameters to laboratory cast
specimens of a given diameter. The effects of size on CP-0.15 and FRP-
0.23 were found by comparing lab made specimens of two different
diameters that were identically cured. Some of the data utilized in this
section were also used in Sections 3.1 or 3.2.
A general rule for producing specimens containing fibers is that the

specimens’ diameter must be at least three times the fiber length. For the
fibers used in this project, that equates to 4- by 8-in. cylinders. Since 3- by
6-in. cylinders were mostly used herein to facilitate comparable testing
with a range of constituents, 4- by 8-in. cylinders were made with fibers
for comparative purposes. With regard to CP, most work to date in this
area has used 2- by 4-in. specimens. Since 3- by 6-in. cylinders were
mostly used herein to facilitate comparable testing with a range of
constituents, 2- by 4-in. cylinders were made with CP for comparative
purposes. In most cases, the compared cylinders were produced from the
same batch for direct comparison.
3.3.1 Size effects – Ultra-High Performance Concrete (UHPC)
Minimal size effects data were collected for UHPC. The only assessment
available was comparing 3- by 6-in. cores to 4- by 8-in. cores where the
cores were taken from the same tub. As such, fiber orientation was
established prior to coring and was comparable in both specimen sizes.
The six cores that were 3- by 6-in. had an average tensile strength of
19.8 MPa, while the three cores that were 4- by 8-in. had an average
tensile strength of 17.0 MPa.
3.3.2 Size effects – Cement paste (CP)
CP size effects data are summarized in Table 10, while Figure 19 plots the
24 individual specimens tested for compressive strength. All specimens
were cured to 2,783 oC-d in the 23 oC curing room.
Of the 24 specimens utilized, six were from Section 3.2 curing effects that
are repeated here and referred to as non-paired, since they were not made
in the same batch as different sized specimens. The remaining
18 specimens were produced in three batches (six specimens per batch),
where one batch produced three specimens that were 2 by 4 in. and three
more specimens that were 3 by 6 in. These specimens could be viewed as
matched pairs.
The paired t-test was performed on the nine matched pairs that could be
produced from batches 1 to 3. The measured compressive strengths for
each batch and specimen size were sorted in ascending order to make the
pairs. For example, the lowest compressive strength of a 2- by 4-in.
specimen from batch 1 was paired with the lowest compressive strength
from a 3- by 6-in. cylinder from batch 1. The comparison was whether the
mean difference of these pairs was statistically different from 0 at a
5 percent level of significance. The p-value was 0.33, indicating the values
were not significantly different. The average strength of the 3- by 6-in.
specimens from batches 1 to 3 was 124 MPa, while the average strength of
the 2- by 4-in. specimens was 119 MPa.
If all 3- by 6-in. specimens were averaged (non-pairs and batches 1 to 3),

the average strength was 118 MPa, which is within 1 MPa of the 2- by 4-in.
specimens. When all data were considered, the standard deviation of the
3- by 6-in. specimens was 17 MPa, while the standard deviation for the 2-
by 4-in. specimens was 21 MPa. Overall, there were no obvious differences
between strengths produced with 2- by 4-in. specimens and those
produced with 3- by 6-in. specimens.
Table 10. Size effects on CP data.

fc fc fc Density
Density Range Density
Mix ID Size Batch Range COV Avg. COV
(g/cm3) Avg. (g/cm3)
(MPa) (%) (MPa) (%)
CP-0.15 3x6 Non-Paired 78.0 – 143.4 21.0 108.7 2.188 – 2.271 1.8 2.232
CP-0.15 2x4 1 94.6 – 144.0 20.7 119.5 2.241 – 2.246 0.1 2.244
CP-0.15 3x6 1 127.3 – 129.9 1.1 128.4 2.247 – 2.254 0.2 2.250
CP-0.15 2x4 2 121.6 – 143.8 9.3 129.9 2.270 – 2.295 0.6 2.284
CP-0.15 3x6 2 119.9 – 138.7 7.4 130.5 2.248 – 2.267 0.4 2.256
CP-0.15 2x4 3 84.7 – 131.9 22.4 106.5 2.251 – 2.263 0.3 2.255
CP-0.15 3x6 3 111.9 – 116.0 1.9 114.4 2.209 – 2.230 0.5 2.222
--Notes: non-paired specimens – 6 fc, 6 densities
Batches 1, 2, and 3 had 3 fc and 3 densities for each specimen size.
Figure 19. Compressive strength of CP-0.15 size variation specimens.
3.3.3 Size effects – Fiber Reinforced Paste (FRP)
FRP size effects data are summarized in Tables 11 (compression) and

12 (tension), while Figures 20 (compression) and 21 (tension) plot the
27 individual specimens tested. All specimens were cured to 2,783 oC-d in
the 23 oC curing room.
Of the 27 specimens utilized, 12 were tested in compression, and 14 were

tested in tension. In compression, six were non-paired (as described in
Section 3.3.2), and six were produced in two batches of three specimens
each. Each batch produced two specimens that were 3 by 6 in. and one
specimen that was 4 by 8 in. In tension, five were non-paired, and nine
were produced in three batches of three specimens each. Each batch
produced two specimens that were 3 by 6 in. and one specimen that was
4 by 8 in.
In compression, the 10 specimens that were 3 by 6 in. had an average

compressive strength of 124 MPa with a range of 114 to 136 MPa. The two
specimens that were 4 by 8 in. had strengths that were 119 and 135 MPa.
Two observations can be made from these data, which are shown in Table
11 and Figure 20. First, the 4- by 8-in. specimens fell in the range of the 3-
by 6-in. specimens. Second, for each batch from which multiple specimens
were made, one of the 3- by 6-in. specimens was essentially the same as
one of the 4- by 8-in. specimens.
Table 11. Size effects of FRP-0.23 in compression.

fc fc fc Density Density Density
Mix ID Size Batch Range COV Avg. Range COV Avg.
(MPa) (%) (MPa) (g/cm3) (%) (g/cm3)
FRP-0.23 3x6 Non-Paired 113.9 – 135.8 6.6 122.2 2.457 – 2.623 1.8 2.531
FRP-0.23 3x6 1 119.1 – 135.9 9.3 127.5 2.527 – 2.538 0.3 2.532
FRP-0.23 4x8 1 134.6 ----- ----- 2.541 ----- -----
FRP-0.23 3x6 2 119.1 – 128.5 5.3 123.8 2.489 – 2.512 0.6 2.500
FRP-0.23 4x8 2 118.9 ----- ----- 2.560 ----- -----
--Note: non-paired specimens – 6 fc, 6 densities
3x6 batched specimens – 2 fc, 2 densities
4x8 batched specimens – 1 fc, 1 density
Figure 20. Compressive strength of FRP variation specimens.
In tension, the 11 specimens that were 3 by 6 in. had an average tensile

strength of 22.2 MPa with a range of 17.7 to 27.6 MPa. The three specimens
that were 4 by 8 in. had strengths that were 17.5, 18.4, and 20.4 MPa with
an average value of 18.8 MPa. Three observations can be made from these
data, which are shown in Table 12 and Figure 21. First, the 4- by 8-in.
specimens fell on the lower end of, to slightly below, the 3- by 6-in.
specimens’ range. Second, the average strength of 4- by 8-in. specimens
was 85 percent of that from the 3- by 6-in. specimens. Third, the 4- by 8-in.
specimen always had the lowest tensile strength in a given batch.
Table 12. Size effects on FRP-0.23 in tension.

St St St Density Density Density
Mix ID Size Batch Range COV Avg. Range COV Avg.
(MPa) (%) (MPa) (g/cm3) (%) (g/cm3)
FRP-0.23 3x6 Non-Paired 17.7 – 27.7 17.8 22.0 2.457 – 2.545 1.3 2.511
FRP-0.23 3x6 3 20.0 – 22.9 9.8 21.5 2.485 – 2.519 1.0 2.502
FRP-0.23 4x8 3 17.5 ----- ----- 2.560 ----- -----
FRP-0.23 3x6 4 21.9 – 22.6 2.2 22.2 2.497 – 2.513 0.5 2.505
FRP-0.23 4x8 4 20.4 ----- ----- 2.565 ----- -----
FRP-0.23 3x6 5 22.3 – 25.0 8.0 23.6 2.519 – 2.545 0.7 2.532
FRP-0.23 4x8 5 18.4 ----- ----- 2.567 ----- -----
--Note: non-paired specimens – 5 st, 5 densities
3x6 batched specimens – 2 st, 2 densities
4x8 batched specimens – 1 st, 1 density
Figure 21. Tensile strength of FRP variation specimens.
3.4 Instrumentation test results

Of the 254 specimens produced, eight were instrumented with strain gages
after 791 oC-d curing. Each specimen was fitted with four strain gages, as
shown in Appendix A. Of the 32 total strain gages, 27 measured data
successfully. As noted earlier in the report, strain readings were taken with a
data acquisition system independent of the concrete compression machine:
load versus time was measured with the compression machine, and strain
versus time was measured with the data acquisition system. These two time
scales were synchronized within, at most, a few seconds so that stress versus
strain plots could be generated. The synchronization process was

approximate but was reasonable for the purposes of this report.
In compression UHPC, CP-0.15, M-0.56, and FRP-0.23 were produced

without replication where three strain gages were oriented vertically
(parallel to axis of loading) and one gage was oriented horizontally.
Figure 22 plots all strain readings taken in compression relative to the
specimen’s compressive strength (fc), which was determined by taking the
stress at any instant in time, dividing by fc, and converting to a percentage.
Strain readings were plotted for all gages up to 100 percent, and thereafter
the curves were visually examined. Stress versus strain plots were truncated
when visual evidence showed gage failure (e.g., a 1,000 microstrain change
in strain from one reading to the next is likely gage debonding and not
specimen response).
In tension UHPC, CP-0.15, M-0.56, and FRP-0.23 were produced without

replication where three strain gages were oriented horizontally
(perpendicular to axis of loading) and one gage was oriented vertically.
Figure 23 plots strain readings relative to the specimen’s tensile strength
(St) in the same manner as Figure 22.
Figure 22. Percent of ultimate compressive stress versus strain for

(a) UHPC, (b) CP-0.15, (c) M-0.56, and (d) FRP-0.23.
(a)
(b))
(c))
(d)
Figure 23. Percent of ultimate tensile stress versus strain for (a)
UHPC, (b) CP-0.15, (c) M-0.56, and (d) FRP-0.23.
(a)
(b)
(c)
(d)
Numerical modeling of UHPC (or any of its constituents) benefits from

knowing how high stresses can get before material behavioral changes are
documented. When unconfined and tested as modest load rates, concrete
materials are generally linear in their stress-strain behavior. A question,
though, is how high can stresses get and linearity remain? As seen in
Figures 22 and 23, linearity was often maintained at stresses nearing
ultimate values, especially in compression. In tension, behaviors were
somewhat more erratic.
Table 13 summarizes the highest ultimate tensile or compressive strain

(εmax) recorded by each gage in Figures 22 and 23 that was deemed
reasonable from the aforementioned truncation process. Alongside this
strain value is the percent of ultimate stress (%UltS) where this strain was
recorded. For compressive loadings, linear behavior until 90 percent or
more of %UltS was generally observed. Note that sampling rates were
relatively low, and no useful information was collected as specimens
approached failure.
Table 13. Strain gage data for compression and tension.

SG-0 SG-1 SG-2 SG-3
Mix Mode εmax %UltS εmax %UltS εmax %UltS εmax %UltS
Compression -3,807 96% --- --- --- --- -3,996 90%
UHPC
tension -973 97% -83 97% 3,381 100% 117 85%
Compression -2,307 98% -2,897 98% 475 98% -3,210 98%
CP-0.15
tension -639 65% 100 36% --- --- 1,315 91%
Compression -4,331 100% -4,393 100% 600 99% -4,426 100%
M-0.56
tension -644 41% 188 41% 311 41% 129 41%
Compression --- --- -5,604 89% 831 89% --- ---
FRP-0.23
tension -1,649 100% 63 100% 5,472 100% -678 100%
The linear portions of Figure 22 (vertical gages only) were used to produce
stress-strain plots up to approximately 40 percent of ultimate stress for
elastic modulus determination (Figure 24); elastic moduli calculated from
strain gage measurements were denoted ESG-0 to ESG-3. The ASTM C469
(ASTM International 2014) compressometer was also fitted to these same
specimens to measure E469 so a direct comparison of strain gage and
compressometer values could be made on the same specimen. Results are
summarized in Table 14.
Figure 24. Truncated stress versus vertical strain plots for (a) UHPC, (b) CP-0.15, (c) M-0.56,
and (d) FRP-0.23 in compression.
70
Note: SG-1 did not
60 record properly.
50
Stress (MPa)
40
(a)
30
C-469
20
SG-0
10
SG-3
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
ɸ*106 (mm/mm)
(b))
70
60 (c))
50
Stress (MPa)
40
30 C-469
20 SG-0
SG-1
10
SG-3
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
ɸ*106 (mm/mm)
(d))
Table 14. Comparison of elastic moduli from strain gages and compressometer.
Mix ID ESG-0 (MPa) ESG-1 (MPa) ESG-3 (MPa) ESG-AVG (MPa) E469 (MPa)
UHPC 42,100 ---- 41,100 41,600 49,610
CP-0.15 43,700 34,200 30,300 36,000 33,800
M-0.56 40,700 40,000 40,300 40,300 45,300
FRP-0.23 ---- 26,800 ---- 26,800 31,400
--Note: All measurements shown in a given row were made on the same specimen.
In three cases, E469 resulted in higher values than strain gages (12 to 19
percent higher on average), and in 1 case, E469 resulted in lower values
than strain gages (7 percent lower on average). Overall, strain gages and
the compressometer reported elastic modulus values that were in the same
general range based on this fairly limited data set.
From compression testing with strain gages, Poisson’s ratio Ǒ was

approximated by relating the horizontal and vertical displacement using
SG-2 (horizontally oriented) and SG-1 (vertically oriented). Figure 25
provides the plots used, which are truncations of Figure 22, and the
corresponding Poisson’s ratio values.
Figure 25. Poisson’s ratio results from compression testing.
3.5 Volume fractions estimated from batch quantities

Table 15 shows the results of the volume fraction estimates from batch
quantities and specific gravities of individual ingredients. All specimens
tested except for cores are represented in Table 15. These fractions are
approximations used for general reference and for comparison with the
ImageJ and Abaqus results presented in the following section.
Table 15. Average volume fractions estimated from batch quantities.

Vc Vsf Vsfl Vs VW Vam Va VTOTAL
Mix Vf (%)
(%) (%) (%) (%) (%) (%) (%) (%)
UHPC 25.3 13.8 8.3 29.1 16.7 3.2 1.3 2.4 100
CP-0.15 44.0 24.0 ----- ----- 29.0 ----- 2.2 0.7 100
M-0.56 30.2 16.5 9.9 18.2 19.8 ----- 1.5 3.9 100
FRP-0.11 41.7 22.7 ----- ----- 27.5 5.2 2.1 0.9 100
CP-0.11 46.0 25.1 ----- ----- 22.8 ----- 2.3 3.8 100
CP-0.26 41.3 22.6 ----- ----- 34.1 ----- 2.1 0.01 100
M-0.47 31.5 17.2 10.4 14.3 20.8 ----- 1.6 4.2 100
M-0.65 28.5 15.6 9.4 21.5 18.8 ----- 1.4 4.8 100
FRP-0.17 42.4 23.2 ----- ----- 28.0 4.0 2.1 0.3 100
FRP-0.23 42.7 23.3 ----- ----- 28.2 2.7 2.1 1.0 100
3.6 Volume fractions estimated from imaging

Of the 254 specimens produced, five specimens (all cured for 791 oC-d)
were allocated for image testing. Four of these specimens (UHPC, CP-0.15,
M-0.56, and FRP-0.23) underwent imaging with an SEM that resulted in
two dimensional (2-D) pictures, and the fifth specimen (UHPC)
underwent a CT scan to get a 3-D representation of the UHPC.
Area fractions for the UHPC specimen evaluated with SEM techniques
were found using ImageJ software. Table 16 compares the volume
fractions in Table 15 to ImageJ values. It is important to note that the area
fraction of cement found through imaging should not match the batching
volume fraction. ImageJ found unhydrated cement grains, while the
batching measures all cement put into the mixture.
One specimen also underwent a CT scan (Figure 26). Large voids, which
can occur during mixing, were observed. Fiber orientation was assumed to
be random, but the CT scan aided in observing the locations of the fibers
within the matrix. The scan found that fibers were 3 percent of the total
volume, which is close to what was calculated from batch quantities and
found using imaging area.
Table 16. Comparison of average UHPC volume

fractions from batching and ImageJ.
Estimated from batching
quantities ImageJ
Inclusion (%) (%)
Air voids 2.4 6.8
Cement 25.3 11.7
Sand 29.1 22.3
Fibers 3.2 3.7
Figure 26. CT scan of UHPC showing (a) approximately 3% steel

fibers and (b) large voids illustrated.
4 Summary
In simple terms, concrete properties are governed by aggregate properties,
paste properties, and the bond between paste and aggregates when
constituent materials and proportions are constant. However, under-
standing the true nature of how concrete properties are governed by these
items is much more complex (even for constant materials). To provide some
fundamental data on Cor-Tuf UHPC, a series of fundamental experiments
was performed and is documented in this report, the ultimate purpose of
which is to further numerical modeling efforts.
Cor-Tuf has seven constituent ingredients that lead to eight volume

fractions when air voids are included. Ten mixtures were produced by
using varying amounts of these seven ingredients and were exposed to
varying curing times and temperatures. These mixtures can be divided
into UHPC (all seven ingredients), FRP (no aggregates), M (no fibers), and
CP (no aggregates or fibers). Mechanical property testing at low load rates
and without confinement measured compressive strength, tensile
strength, elastic modulus, Poisson’s ratio, and load versus strain
relationships. The majority of the specimens produced were 3 in. in
diameter and had a height of 6 in., since this is a mid-range size for the
types of mixtures produced. Some specimens were 2 in. by 4 in. or 4 in. by
8 in. to assess size effects.
This report is intended primarily to document these experiments and

contain the data collected, since a systematic evaluation of this nature
related to constituent effects is not readily available to the knowledge of
the authors. Specific conclusions are avoided herein as the intent is to use
these data in future efforts that will be more appropriate to draw more
meaningful conclusions about ways to better model and ultimately
improve UHPC. An intended purpose for these data is high-performance
computing (HPC) environments where multi-scale modeling is performed.
Fundamental understanding of individual components, proportions, and
their interface behavior is needed for sophisticated numerical models
capable of predicting behavior of structural systems built with UHPC. The
experimental data collected have the possibility to be used three ways for
numerical modeling purposes: (1) exploratory efforts to assist in defining
first principles, (2) calibration, and (3) validation.
The following list summarizes key observations from this report.
1. Instrumented compression specimens showed mostly linear behavior

to stress levels of 90 percent (or more) of failure levels.
2. Specimens produced with UHPC constituent materials did not produce
as high of an elastic modulus per unit of compressive strength as does
typical ready mixed concrete.
3. With CP, there were no overall obvious compressive strength
differences between 2- by 4-in. specimens and 3- by 6-in. specimens.
4. With FRP, compressive strengths from 4- by 8-in. specimens fell in the
range of those from 3- by 6-in. specimens. Tensile strengths from 4- by
8-in. specimens fell on the lower end of to slightly lower than 3- by
6-in. specimens. Note that 4- by 8-in. specimen data were limited.
5. Specific aspects of curing seem to affect compressive and tensile
properties, and the effects are not consistent between tension and
compression or between types of specimen. These data suggest that
only accumulating oC-days and using that number as a maturity index
may lead to undesirable outcomes for some types of UHPC endeavors,
such as mass concrete.
References
Abaqus CAE (version 6.14). 2014. Dassault Systems.
American Concrete Institute. 2014. Building code requirements for structural concrete
(ACI 318-05) and commentary (ACI 318R-05). Farmington Hills, MI: American
Concrete Institute.
American Petroleum Institute. 1991. Specification for well cements, API specification
IOA. 21st ed. September, 1991. Washington DC: American Petroleum Institute.
ASTM International (ASTM). 2011a. Standard practice for estimating concrete strength
by the maturity method. Designation C1074-11. West Conshohocken, PA: ASTM
International. Doi: 10.1520/C1074-11. www.astm.org/
_____. 2011b. Standard test method for splitting tensile strength of cylindrical concrete
specimens. Designation C496/C496M. West Conshohocken, PA: ASTM
International. Doi 10.1520/C0496_Co496M-11. https://www.astm.org/
_____. 2013. Standard specification for chemical admixtures for use in producing
flowing concrete. Designation C1017/C1017M-13E01. West Conshohocken, PA:
ASTM International. Doi 10.1520/C1017_C 1017M-13E01. https://www.astm.org/
_____. 2014. Standard test method for static modulus of elasticity and Poisson’s ration
of concrete in compression. Designation: C469/C469M. West Conshohocken,
PA: ASTM International. Doi: 10.1520/C0469M-14. https://www.astm.org/
_____. 2015. Standard test methods for chemical analysis of hydraulic cement.
Designation: C114/C114M. West Conshohocken, PA: ASTM International. Doi:
10.1520/C0114-15. https://www.astm.org/
_____. 2016a. Standard specification for steel fibers for fiber-reinforced concrete. West
Conshohocken, PA: ASTM International. Doi: 10.1520/A0820_A0820M-16.
https://www.astm.org/
_____. 2016b. Standard test method for compressive strength of cylindrical concrete
specimens. Designation: C39/C39M. West Conshohocken, PA: ASTM
International. Doi: 10.1520/C0039_C0039M-16. https://www.astm.org/
_____. 2017a. Standard specification for chemical admixtures for concrete.

Designation: C494/C494M. West Conshohocken, PA: ASTM International. Doi:
10.1520/C0494_C0494M-17. https://www.astm.org/
_____. 2017b. Standard test methods for fineness of hydraulic cement by air-
permeability apparatus. Designation: C204/C204M. West Conshohocken, PA:
ASTM International. Doi: 10.1520 C0204-17. https://www.astm.org/
Chandler, M., J. Peters, and D. Pelessone. 2012. Modeling nanomechanical behavior of

calcium-silicate-hydrate. ERDC/GSL TR-12-30. Vicksburg, MS: U.S. Army
Engineer and Research and Development Center.
Green, B., R. Moser, D. Scott, and W. Long. 2014. Ultra-high performance concrete
history and usage by the Corps of Engineers. Advances in Civil Engineering
Materials 4(2): 132-143. https://doi.org/10.1520/ACEM20140031.
Howard, I. L., and H. V. GangaRao. 2009. Testing and evaluation of adhesive bonded
fiber reinforced polymer bridge decks. Journal of Advanced Materials 41(2): 28-
46.
Roth, M. J. 2008. Flexural and tensile properties of thin, very high-strength, fiber-
reinforced concrete panels. ERDC/GSL TR-08-24. Vicksburg, MS: U.S. Army
Engineer Research and Development Center.
Rushing, T. W., I. L. Howard, J. B. Jordan, and P. G. Allison. 2016. Laboratory

characterization of fatigue performance of AM2 aluminum airfield matting.
Journal of Materials in Civil Engineering 28(11): 04016134.
Schneider, C. A., W. S. Rasband, and K. W. Eliceiri. 2012. NIH image to ImageJ: 25 years
of image analysis. Nature Methods 9(7): 671–75.
Scott, D., W. Long, R. Moser, B. Green, J. O’Daniel, and B. Williams. 2015. Impact of steel
fiber size and shape on the mechanical properties of ultra-high performance
concrete. ERDC/GSL TR-15-22. Vicksburg, MS: U.S. Army Engineer Research
and Development Center.
Warren, K. A., B. Christopher, and I. L. Howard. 2010. Geosynthetic strain gage

installation procedures and alternative strain measurement methods for roadway
applications. Geosynthetics International 17(6): 403-430.
Appendix A: Strain Gage Locations

Figure A1. Strain gage locations of UHPC in compression.
Figure A2. Strain gage locations for CP in compression.

Figure A3. Strain gage locations for M in compression.

Figure A4. Strain gage locations for FRP in compression.

Figure A5. Strain gage locations for UHPC in tension.

Figure A6. Strain gage locations for CP in tension.

Figure A7. Strain gage locations for M in tension.

Figure A8. Strain gage locations for FRP in tension.

Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining
the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for
reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA
22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display
a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To)
October 2018 Final report
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Mechanical Behavior of Cor-Tuf Ultra-High-Performance Concrete Considering Aggregate and Paste W56HZV-17-C-0095
Effects 5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER

Isaac L. Howard, Ashley Carey, Megan Burcham, Dylan A. Scott, Jameson D. Shannon, Robert D. 458161
Moser, and Mark F. Horstemeyer 5e. TASK NUMBER
5f. WORK UNIT NUMBER

J495LJ
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT
NUMBER
U.S. Army Engineer Research and Development Center
Geotechnical and Structures Laboratory ERDC/GSL TR-18-31
3909 Halls Ferry Road
Vicksburg, MS 39180-6199
9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

U.S. Army Corps of Engineers
Washington DC 20314-1000
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION / AVAILABILITY STATEMENT
13. SUPPLEMENTARY NOTES
14. ABSTRACT
This research primarily focused on properties from varying the constituents that make up ultra-high performance concrete (UHPC) with the ultimate
goal to enable improved characterization and modeling of this material. Several variations of UHPC were made to see the differences in properties as a
function of constituents. Compressive strength, elastic modulus, and tensile strength were measured at low loading rates. Fundamental test methods
were used for most experiments with a smaller subset of tests with strain gages and imaging techniques. This report is intended primarily to document
these experiments and the collected data. Specific conclusions are avoided herein, as the intent is to use these data in future efforts that will be more
appropriate to draw more meaningful conclusions about ways to better model and ultimately improve UHPC.
15. SUBJECT TERMS Concrete-Mixing, Concrete-Additives, Aggregates (Building materials)

UHPC, Cor-Tuf, Concrete-Mechanical High strength concrete,
properties, Characterization, microCT, Modeling, Concrete,
16. SECURITY CLASSIFICATION OF: 17. LIMITATION 18. NUMBER 19a. NAME OF RESPONSIBLE
OF ABSTRACT OF PAGES PERSON Robert D. Moser
a. REPORT b. ABSTRACT c. THIS PAGE 19b. TELEPHONE NUMBER (include
area code)
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED 74
601.634.3261
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std. 239.18

Mechanical Behavior of Cor Tuf Ultra Hig

Uploaded by

Copyright:

Available Formats

Mechanical Behavior of Cor Tuf Ultra Hig

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mechanical Behavior of Cor Tuf Ultra Hig

Uploaded by

Copyright:

Available Formats

ERDC/GSL TR-18-31

Mechanical Behavior of Cor-Tuf Ultra-High

Approved for public release; distribution is unlimited.

Mechanical Behavior of Cor-Tuf Ultra-High

Ashley Carey, Megan Burcham, and Mark F. Horstemeyer

Prepared for U.S. Army Corps of Engineers

Figures and Tables .........................................................................................................................................v

Unit Conversion Factors ........................................................................................................................... viii

2 Experimental Program ......................................................................................................................... 7

3 Test Results ..........................................................................................................................................25

Appendix A: Strain Gage Locations .........................................................................................................56

Report Documentation Page

Figures and Tables

Figure A5. Strain gage locations for UHPC in tension. ............................................................................60

Unit Conversion Factors

UHPC is usually composed of five to eight individual materials with

ERDC. An ultimate goal is for this report to be used alongside additional

This report represents work performed beginning in 2013 under three

1.4 Symbols and acronyms

ACI – American Concrete Institute

API – American Petroleum Institute

ASTM – American Society for Testing and Materials

C469 – elastic modulus constant found from the relationship of

CAVS – Center for Advanced Vehicular Systems

CMRC – Construction Materials and Research Center

COV – coefficient of variation, standard deviation divided by mean

CP-0.11 – cement paste with a water to cementitious material ratio of 0.11

CP-0.15 – cement paste with a water to cementitious material ratio of 0.15

CP-0.26 – cement paste with a water to cementitious material ratio of 0.26

E469 – elastic modulus from ASTM C469

ESG – elastic modulus measured by using strain gages to measure strain

ERDC – U.S. Army Engineer Research and Development Center

ERS – Engineered Resilient Systems

FRP – fiber reinforced paste

FRP-0.11 – fiber reinforced paste with fibers to cementitious material ratio

FRP-0.17 – fiber reinforced paste with fibers to cementitious material ratio

FRP-0.23 – fiber reinforced paste with fibers to cementitious material

HPC – High Performance Computing

HSR – High Sulfate Resistance

IVPPED – Integrated Virtual Prototyping for Product Engineering and

M-0.47 – mortar with a fine aggregate to cementitious material ratio of 0.47

M-0.56 – mortar with a fine aggregate to cementitious material ratio of 0.56

M-0.65 – mortar with a fine aggregate to cementitious material ratio of 0.65

MSU – Mississippi State University

oC-d – measure of maturity that is degree Celsius * days

RVE – Representative Volume Element

SEM – Scanning Electron Microscope

UHPC – Ultra-high performance concrete

Vam – volume of admixture

Vnoair – volume of the specimen if there were no air

Vsf – volume of silica fume

Vsfl – volume of silica flour

VTotal – total volume

WD63 – Engineering Work Directive 0063

cm – cementitious material (class H cement and silica fume)

ddial – displacement of the compressometer dial

f/cm ratio – ratio of fibers to cementitious materials

fa – fine aggregate (silica flour and silica sand)

fa/cm – ratio of fine aggregate to cementitious materials