PHYSICAL PROPERTIES AND CHARACTERISTICS

AFFECTING THE SENSITIVITY TO CRACKING OF

CEMENTITIOUS REPAIR MATERIALS
Keywords: coefficient of thermal expansion; cracking; creep; dimensional compatibility; drying shrinkage; durability; elastic modulus; phys­
ical properties; repair material; tensile strength.

Question:
What properties and characteristics of cement-based repair materials influence cracking in repairs?

Discussion
One of the main factors assuring the durability and long-term performance of concrete repairs is its resistance
to cracking. Cracks are open pathways providing aggressive agents easy access into the repair, promoting the
development of reinforcement corrosion and deterioration of the repaired structure (Vaysburd 1995; ACI 224.1R).
Repair materials typically crack as a result of restrained volume changes. To achieve dimensional compatibility
and minimize cracking in repairs, the relevant physical properties and characteristics of the repair material
should be appraised during the selection process, notably in the case of cement-based repair materials. The
cracking sensitivity of concrete repairs is also influenced by factors such as surface preparation; patch geom­
etry; and presence of reinforcing steel, placement conditions (notably the temperature and relative humidity of
both the existing concrete and ambient air, which may result in significant gradients within the repair), curing,
and expansive forces in the existing concrete, which are addressed elsewhere and not discussed herein.

Structural and nonstructural repairs are considered in this discussion. Structural repairs are intended to
increase the load-carrying capacity of a structural component beyond its current capacity or to restore a
damaged structural component to its original design load-carrying capacity. For example, repairs to structural
elements such as columns subjected to applied loads should accommodate these loads. Conversely, protective
repairs performed to reestablish the original configuration without altering the structural capacity of a member
are generally defined as nonstructural. Nevertheless, in ACI 562, the definition of a structural repair has been
widened to include any repair that creates an unsafe condition in the event it fails, irrespective of the structural
capacity consideration.

Cracking caused by restrained contraction occurs when the induced tensile stress exceeds the tensile strength of
the repair material. External and internal restraint conditions that induce tensile stresses that can lead to cracking
are discussed in ACI 207.2R and ACI 224.1R. Restrained contractions that can induce cracking are primarily caused
by shrinkage and thermal volume changes. In the case of cementitious repair materials, these volume changes may
occur individually or in combination, while the material is in either a plastic or hardened state.

The magnitude of shrinkage is generally the dominant factor, but not the only factor affecting the cracking
sensitivity of a repair material. The other important factors and characteristics interplaying in the overall
cracking behavior of the material are (Vaysburd et a!. 1999, 2000, 2014; McDonald et a!. 2002; Bissonnette et a!.
2015; Courard et a!. 2015):
a) Degree of restraint
b) Tensile strength
c) Modulus of elasticity
d) Creep
e) Coefficient of thermal expansion

The combination of properties that are most desirable to reduce the advent of cracking in cement-based
materials can be lumped into the single notion of extensibility (Mehta and Monteiro 1993), which corresponds
2 CHARACTERISTICS AFFECTING THE SENSITIVITY TO CRACKING OF CEMENTITIOUS REPAIR MATERIALS (ACI364.16T-18)

the sum of their short-term (elastic) and long-term (creep) tensile strain capacity. Cement-based materials with
a high degree of extensibility can resist volume changes (deformations) without cracking. Therefore, to limit
the potential for cracking, cement-based repair materials should have low shrinkage, and should also exhibit as
much extensibility as possible, through a combination of low elastic modulus, high creep, high tensile strength,
or all these.

Overall, the requirement for long-lasting monolithic behavior is that the repair materials have properties and
dimensional behavior that will make them compatible with the existing concrete substrate for the application
considered. Dimensional compatibility is defined as a balance of strains between a repair material and the
existing substrate, such that the composite repair system withstands all stresses induced by the various volume
changes without distress and deterioration over a designed period of time.

A comprehensive investigation conducted by the U.S. Army Corps of Engineers (Vaysburd et a!. 1999; McDonald
et a!. 2002) evaluated the relationship between these properties, characteristics, and the repair material's field
performance with respect to crack resistance. While the results did not reveal conclusive trends when consid­
ering the properties and characteristics individually, a correlation with the cracking sensitivity was found when
considering them together. The in-depth analysis of the data generated in both laboratory and field experiments
led to the identification of preliminary performance criteria for the most influential repair material properties
and characteristics. The relative importance of the latter varies depending on application and service condi­
tions; therefore, the requirements should be modified as necessary to achieve compatibility with the existing
substrate.

Tensile strength Intuitively, the easiest way to improve the resistance of cementitious materials to cracking
-

would be to achieve substantially high tensile strength. In reality, because the tensile capacity of cement-based
materials cannot be increased substantially, the primary goal should be to reduce tensile stresses to minimize
cracking. Nonetheless, it was concluded in the study by the U.S. Army Corps of Engineers (Vaysburd et a!. 1999;
McDonald et a!. 2002) that the tensile strength of the repair material should be a minimum of 400 psi (2.8 MPa).
Such a value would be expected for conventional concrete with a compressive strength of approximately 4000
psi (28 MPa).

The term "tensile strength" has no absolute meaning for cement-based materials and needs to be expressed
in terms of the specific test procedure used. Three types of tests are primarily used for cementitious materials
tension testing: the direct tension test, the flexural test, and the splitting tensile test. The direct tension test
is challenging to perform because of the inherent difficulty of ensuring that the load is truly axial. In a ductile
material, some eccentricity will not have much effect on tensile strength, but in brittle cementitious mate­
rials, there is relatively little stress redistribution and, consequently, the test gives an underestimate of true
tensile strength. Still, when properly conducted, a direct tension test such as the one described in the USACoE
CRD-C 164 procedure yields a more representative measure than indirect methods. The results from the flexural
strength or beam test (modulus of rupture) and the splitting-strength test are both known to overestimate the
tensile strength of the material. Data by Price (1951) demonstrate that flexural strength testing tends to over­
estimate the tensile strength of concrete by 50 to 100 percent. Despite their simplicity, the indirect tests fail to
reproduce the state of stress developed within a centrically tensioned specimen and, consequently, the values
of the corresponding strengths differ from the pure tensile strength values.

The compressive strength of a repair material intended for structural repair should ideally be similar to that of
the existing concrete. Generally, it is not desirable to have repair materials with compressive strengths in excess
of 4000 psi (28 MPa) for nonstructural repairs. Rather, it is generally agreed that high-strength repair materials
are more prone to cracking because of high stresses developed from restrained shrinkage, as a result of typi­
cally high elastic modulus and lower creep. In addition, the use of rapid-setting materials to achieve high early
strength often leads to increased cracking because of higher early-age volume changes, increased stiffness, and
less creep.

Modulus of elasticity-In a repair, the modulus of elasticity controls, to some extent, the load sharing between
the repaired area and the rest of the element, as well as the level of tensile stress generated due to volume
changes such as shrinkage and thermal deformations. For structural repair, the modulus of elasticity of the
repair material should be compatible with that of the existing concrete. Conversely, it was concluded in the
USACoE CRD-C 164 investigation that the modulus of elasticity of materials for nonstructural repairs, deter­
mined in accordance with ASTM C469/C469M, should be specified not to exceed 3.5 x 106 psi (24 GPa).

American Concrete Institute- Copyrighted© Material- www.concrete.org

 CHARACTERISTICS AFFECTING THE SENSITIVITY TO CRACKING OF CEMENTITIOUS REPAIR MATERIALS (ACI364.16T-18) 3

Such a low limit might appear somewhat discriminant in view of the range of values typically found for
cementitious materials, especially considering the recommended minimum tensile strength of 400 psi (2.8 MPa).
What should primarily be seen in that recommendation is the need for the lowest effective material stiffness to
minimize the restrained shrinkage stresses and the highest extensibility to prevent shrinkage-induced cracking.
The effective stiffness (ACI 209R), takes into account both the short-term (elastic modulus) and the long-term
(creep) deformational behavior of the material. The rationale of this approach is that shrinkage is a time-depen­
dent phenomenon and restrained shrinkage stresses are induced progressively within the repair material, thus
leaving time for creep and stress relaxation to occur.

Nevertheless, lowering the repair material stiffness is very desirable in any nonstructural repairs. The factors
affecting the modulus of elasticity of cement-based materials are related to compressive strength and density.
Hence, factors that affect strength such as cement content, wjcm, aggregate size, type and grading, curing condi­
tions, and age at the time of testing should similarly influence modulus. The modulus of elasticity of cement­
based materials can be reduced with a higher wjcm and aggregate with a lower modulus of elasticity.

Cracks exist in a composite repair system for reasons other than service loads. Some of these are caused not only
by the drying shrinkage of the repair material and the difference in the coefficients of thermal expansion from the
substrate material, but also to the differences in the elastic moduli of the repair material and substrate concrete.

The compatibility in elastic moduli, therefore, becomes, in cases of structural repairs, an important factor
because incompatibility may lead to considerable stress concentration when widely differential volume changes
of the repair material in relation to the concrete substrate occur. Because, in such situations, the interfacial
bond region (transition zone) is the weak link in the repair system, cracks will tend to form in this region. In
certain cases where bond strength is high, cracks will occur in the matrix of the material, which has the highest
modulus of elasticity. When external load is perpendicular to the bond line, as in the case of repaired pave­
ment, differences in modulus of elasticity regarding the varying stiffness between repair materials and concrete
substrate is normally not problematic (Vaysburd et a!. 2014).

Conversely, in repairs where the service load is parallel to the bond line, differences in modulus of elasticity
may cause load transfer to the high-modulus material if the other materials yield under the stress. If the load
transfer is beyond the load-bearing capacity of the higher-modulus material, it will fracture and damage the
structure.

Creep-Relaxation through tensile creep reduces the stresses induced in a repair system by restrained drying
shrinkage, thereby enhancing the crack resistance of repairs. Test results show tensile creep capacity is depen­
dent on material composition, often more so than shrinkage (Pigeon and Bissonnette 1999). Because a reduction
in paste content reduces shrinkage and appears to increase tensile creep, a proper repair mixture should have
the lowest practical cement content. Also, in general, creep of cement-based materials is inversely proportional
to the modulus of elasticity and compressive strength; therefore, high-strength, high-modulus materials are
generally not desirable for nonstructural concrete repairs.

The phenomenon of gradual increase in strain with time under a given level of sustained stress is called
"creep". In the same fashion, the phenomenon of gradual decrease in stress with time under a given level of
sustained strain is called "stress relaxation". Both manifestations are typical of viscoelastic materials. When the
volume changes of a cement-based repair material are restrained, their viscoelastic behavior will translate into
a lower stress-increase rate. Thus, under the restraining conditions present in the repair material, the interplay
between elastic tensile stresses induced by shrinkage strains and stress relief due to creep is at the heart of the
risk of cracking.

Compressive creep properties are easier to obtain and evaluate, but their use in place of tensile creep to
analyze the extensibility and cracking resistance of cement-based repair materials was found to produce signifi­
cant errors (Vaysburd et a!. 1997). The works by Bissonnette and Pigeon (1995) and Bissonnette et a!. (2005)
have highlighted the criticality of tensile creep among the properties of cement-based materials influencing the
behavior of a concrete repair.

Coefficient of thermal expansion (CTE)-The CTE gives a measure of dimensional contraction or expansion
with changes in temperature, and therefore is an essential property of the repair materials used in composite
repair systems. When significant changes in temperature occur, a marked difference in the CTE between repair
and substrate will produce different volume changes, resulting in excessive stresses at the interface, which may
cause bond failure or failure within the lower strength material of the composite repaired structure.

American Concrete Institute- Copyrighted© Material- www.concrete.org

4 CHARACTERISTICS AFFECTING THE SENSITIVITY TO CRACKING OF CEMENTITIOUS REPAIR MATERIALS (ACI364.16T-18)

Except under extreme temperature conditions, surface concrete repairs generally suffer very little or no
distress from changes in temperature, because the CTE of most cement-based repair materials is very close
to that of the existing concrete. The risk for thermal compatibility is most important in environments that are
frequently subject to large changes in temperature, especially large repairs and overlays. In general, the coef­
ficient of thermal expansion, determined in accordance with USACoE CRD-C 39, should not exceed 7 x I0-6;oF
(13 x 1o-6;oC).

However, in large massive repairs (exceeding 12 in. [300 mm] in thickness) and repairs with some very fast­
setting materials, the combination of increased heat of hydration and relatively slow heat-dissipation conditions
may result in a significant rise in temperature. Subsequently, cooling to the ambient temperatures may cause
the repair to crack. For cement-based materials under such conditions, the restrained contraction strain from
cooling is generally most critical.

Drying shrinkage Drying shrinkage of a repair material is affected by a number of factors, including tempera­
-

ture, relative humidity, absorptivity of the substrate, and ratio of volume-to-exposed surface (ACI 209R). The
free drying shrinkage of repair materials can be determined in accordance with ASTM C157/C157M, modified per
ACI 364.3R. The ultimate drying shrinkage value, determined in accordance with ASTM C596, should not exceed
0.05 and 0.1 percent at 28 days and 1 year, respectively.

Wet curing does not eliminate or significantly reduce drying shrinkage; it merely delays moisture evaporation
and resulting shrinkage. Nevertheless, when adequately performed, it can have a beneficial effect on the devel­
opment of material characteristics, such as extensibility (or tensile strain capacity), that will in turn reduce the
risk of cracking related to restrained shrinkage. The extensibility of repair materials is very sensitive to the rate
at which strain is applied or, in other words, to the rate at which drying occurs. Rapid drying causes a rapid
strain rate and minimizes strain relaxation resulting from creep of the material.

The sensitivity of repair materials to restrained shrinkage cracking is dependent on a variety of factors, as
summarized in Table 1 (Snover et a!. 2011). Restrained shrinkage cracking sensitivity should be evaluated in a
ring test in accordance with ASTM C1581/C1581M.

Finally, constituent materials and mixture proportions can have a significant effect on dimensional stability
and resistance to cracking. When selecting or propor­
tioning cement-based repair materials, the following Table 1-Relative effects of material properties and
general guidelines should be followed (Vaysburd et a!. contents on cracking of repairs (Snover et al. 20II)
2006): Effect
a) Low early strength and early elastic modulus, and
Parameter Major Moderate Minor
moderate compressive strength are desirable properties.
b) The cementitious materials content should be kept Drying shrinkage X

as low as possible. Modulus of elasticity X

c) Supplementary cementing materials, Type II Creep X
cement, or both, should be used.
Compressive strength X
d) The total paste content should be minimized.
e) High-quality aggregates should be used and their Early strength X
volume fraction should be optimized (paste content Paste content X
should be minimized).
Cement content and type X

ACI 224.1R provides additional information on the Aggregate content, type, and size X
causes and control of cracking. Evaluation of the prop­ Coefficient of thermal expansion X
erties and characteristics of concrete discussed herein
Water-cementitious materials
is covered in ACI 546.3R and ICRI 320.2R; both are X
ratio
intended to improve and standardize the information
provided in cement-based repair material data sheets. Accelerating admixtures X

Plasticizers X
Summary Silica fume X
Dimensional compatibility determines the ability of
Fly ash X
repairs to resist cracking. The material properties and
Slag X
characteristics that directly influence the strain balance in
a repair system are tensile strength, modulus of elasticity, Water content X
creep, thermal expansion, and shrinkage. As dimensional Slump (within typical ranges) X

American Concrete Institute- Copyrighted© Material- www.concrete.org

 CHARACTERISTICS AFFECTING THE SENSITIVITY TO CRACKING OF CEMENTITIOUS REPAIR MATERIALS (ACI364.16T-18) 5

incompatibility can adversely affect both the load-carrying capacity and the durability of a repaired element, it needs
to be addressed in structural repairs as well as in nonstructural repairs. This TechNote provides general guidelines
to achieve dimensional compatibility of repairs.

References

American Concrete Institute

ACI 207.2R-07-Report on Thermal and Volume Change Effects on Cracking of Mass Concrete
ACI 209R-92(08)-Prediction of Creep, Shrinkage, and Temperature Effects in Concrete Structures
ACI 224.1R-07-Causes, Evaluation and Repair of Cracks in Concrete Structures
ACI 364.3R-09-Guide for Cementitious Repair Material Data Sheet
ACI 546.3R-14-Guide to Materials Selection for Concrete Repair
ACI 562-16-Code Requirements for Assessment, Repair, and Rehabilitation of Existing Concrete Structures
and Commentary

ASTM International
ASTM C157/C157M-08(2014)-Standard Test Method for Length Change of Hardened Hydraulic-Cement Mortar
and Concrete
ASTM C469/C469M-14-Standard Test Method for Static Modulus of Elasticity and Poisson's Ratio of Concrete
in Compression
ASTM C596-09-Standard Test Method for Drying Shrinkage of Mortar Containing Hydraulic Cement
ASTM C1581/C1581M-09-Standard Test Method for Determining Age at Cracking and Induced Tensile Stress
Characteristics of Mortar and Concrete under Restrained Shrinkage

International Concrete Repair Institute

ICRI 320.2R-2009-Guide for Selecting and Specifying Materials for Repair of Concrete Surfaces

Authored references
Bissonnette, B.; Modjabi-Sangnier, F.; Courard, L.; Garbacz, A.; and Vaysburd, A. M., 2015, "A Quantitative
Approach to the Concept of Concrete Repair Compatibility," Concrete Repair, Rehabilitation and Retrofitting IV, 4th
International Conference on Concrete Repair, Rehabilitation and Retrofitting OCCRRR-4), Leipzig, Germany.
Bissonnette, B., and Pigeon, M., 1995, "Tensile Creep at Early Ages of Ordinary, Silica Fume, and Fiber-Rein­
forced Concretes," Cement and Concrete Research, V. 25, No. 5, pp. 1075-1085. doi: 10.1016/0008-8846(95)00102-1
Bissonnette, B.; Pigeon, M.; and Vaysburd, A. M., 2005, "Tensile Creep of Concrete: Study of its Sensitivity to
Basic Parameters," ACI Materials Journal, V. 104, No. 4, July-Aug., pp. 360-368.
Courard, L.; Bissonnette, B.; and Garbacz, A., 2015, "Fundamental Approach for the Concept of Concrete Repair
Compatibility," Concrete Repair, Rehabilitation and Retrofitting IV, 4th International Conference on Concrete Repair,
Rehabilitation and Retrofitting (ICCRRR-4), Leipzig, Germany.
McDonald, J. E.; Vaysburd, A. M.; Emmons, P. H.; Poston, R. W.; and Kesner, K. E., 2002, "Selecting Durable
Repair Materials: Performance Criteria- Summary," Concrete International, V. 24, No. 1, Jan., pp. 37-44.
Mehta, P. K., and Monteiro, P. J., 1993, Concrete: Structure, Properties and Materials, second edition, Prentice Hall,
548 pp.
Pigeon, M., and Bissonnette, B., 1999, "Tensile Creep & Cracking Potential," Concrete International, V. 21, No. 11,
Nov., pp. 31-35.
Price, W. H., 1951, "Factors Influencing Concrete Strength," ACI Journal Proceedings, V. 47, No. 2, Feb., pp. 417-432.
Snover, R. M.; Vaysburd, A. M.; and Bissonnette, B., 2011, "Concrete Repair Specifications: Guidance or Confu­
sion?" Concrete International, V. 33, No. 12, Dec., pp. 57-63.
Vaysburd, A. M., 1995, "Research Needs for Establishing Material Properties to Minimize Cracking in Concrete
Repairs: Summary of a Workshop," NIST, Gaithersburg, MD, 32 pp.
Vaysburd, A. M.; Bissonnette, B.; and von Fay, K. F., 2014, "Compatibility Issues in Design and Implementation
of Concrete Repair Overlays," Report No. MERL-2014-87, U.S. Bureau of Reclamation, Technical Service Center,
Denver, CO, Dec., 136 pp.
Vaysburd, A. M.; Brown, C. D.; and Bissonnette, B., 2006, ""Roadmap for Improvement of Crack Resistance of
Repair Materials," Applicator, V. 28, No. 2, pp. 18-22.
Vaysburd, A. M.; Carino, N.J.; and Bissonnette, B., 2000, "Predicting the Performance of Concrete Repair Mate­
rials," NISTIR 6402, Summary of Workshop, Durham, NH, Jan., 39 pp.
Vaysburd, A. M.; Emmons, P. H.; McDonald, J. E.; Poston, R. W.; and Kesner, K. E., 1999, "Performance Criteria
for Concrete Repair Materials, Phase II Summary Report," Technical Report REMR-CS-62, U.S. Army Engineer
Waterways Experiment Station, Vicksburg, MS, 72 pp.

American Concrete Institute- Copyrighted© Material- www.concrete.org

6 CHARACTERISTICS AFFECTING THE SENSITIVITY TO CRACKING OF CEMENTITIOUS REPAIR MATERIALS (ACI364.16T-18)

Vaysburd, A. M.; Sabnis, G. M.; and Emmons, P. H., 1997, "Minimizing Cracking- The Key to Durable Concrete
Repair," Indian Concrete Journal, V. 71, No. 12, Dec., pp. 652-657.
USACoE CRD-C 39-81, 1981, 'Test Method for Coefficient of Linear Thermal Expansion of Concrete," Handbook
Concrete and Cement, U.S. Army Engineer Waterways Experiment Station, V icksburg, MS, June, 2 pp.
USACoE CRD-C 164-92, 1992, "Standard Test Method for Direct Tensile Strength of Cylindrical Concrete or
Mortar Specimens," Handbook Concrete and Cement, U.S. Army Engineer Waterways Experiment Station, V icks­
burg, MS, Dec., 4 pp.

Reported by ACI Committee 364

Ashok M. Kakade, Chair Paul E. Gaudette, Secretary

Randal M. Beard LiyingJiang Consulting Members

Benoit Bissonnette
Ryan Alexander Carris
Larry D. Church
Bruce A. Collins
Timothy R. W. Gillespie
Fred R. Goodwin
Pawan R. Gupta
Ann Harrer
John L. Hausfeld
Robert L. Henry
CharlesJ. Hookham
Keith E. Kesner Robert V. Gevecker
John S. Lund Stephen A.Johanson
Marjorie M. Lynch Emory L. Kemp
Surendra K. Manjrekar Howard H. NewlonJr.
James E. McDonald Weilan Song
Jay H. Paul DeJa Tharmabala
Murat B. Seyidoglu Robert Tracy
K. Nam Shiu Alexander M. Vaysburd
Kyle D. Stanish William F. Wescott
David A. VanOcker
David W. Whitmore

American Concrete Institute- Copyrighted© Material- www.concrete.org