Citation
Nondestructive method for hardness evaluation of mortars

Material Information

Title:
Nondestructive method for hardness evaluation of mortars
Creator:
Transue, David John
Publication Date:
Language:
English
Physical Description:
xiii, 110 leaves : illustrations (some color) ; 28 cm

Subjects

Subjects / Keywords:
Mortars -- Testing ( lcsh )
Hardness -- Testing ( lcsh )
Masonry ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 108-110).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by David John Transue.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
42612623 ( OCLC )
ocm42612623
Classification:
LD1190.E53 1999m .T73 ( lcc )

Full Text
NONDESTRUCTIVE METHOD FOR
HARDNESS EVALUATION OF MORTARS
by
David John Transue
B.A., The College of Wooster, 1988
B.S., University of Colorado, 1996
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
1999
i


This thesis for the Master of Science
degree by
David John Transue
has been approved
by
Michael Schuller

Date


Transue, David John (M.S., Civil Engineering)
Nondestructive Method for Hardness Evaluation of Mortars
Thesis directed by Assistant Professor Kevin L. Rens
ABSTRACT
The objective of this study was to develop a test method to measure the hardness of
masonry mortars with a pendulum hammer. The results of this test method are to be
used as an aid in the selection of pointing mortars and for the evaluation and quality
control of in-place mortars. Eight mortar formulations were investigated in the
laboratory. These tests included measures of the plastic properties of the mortars, the
compressive strength, and the water vapor transmission. The laboratory tests were
conducted in accordance with American Society for Testing and Materials (ASTM)
standard test procedures where appropriate. Masonry piers for pendulum hammer
hardness testing were built using the same eight mortar formulations. The results of
mortar hardness measurements from these piers were compared to the properties
determined in the laboratory testing program.
Field tests were conducted on historic masonry structures in Colorado and across the
United States. The experience gained from the field testing and the results of the
laboratory comparisons were used to develop a test method for hardness
determination using the pendulum hammer. This method includes guidelines for use
of the equipment and provides a method for obtaining statistically significant results.
A standard test method for this technique has been drafted and submitted to ASTM
for consideration as a new standard.
A literature review was performed focusing on existing techniques of evaluating
historic and modem mortars, including hardness testing, penetration and pullout
techniques, drilling resistance, and chemical testing. A finite element analysis of
stresses in a brick bedded on two mortars of different elastic moduli was performed.


The results reported herein indicate that the results of pendulum hammer testing
correlate well to the mortar type and compressive strength. The hardness results do
not correlate well with the results of the water vapor transmission tests nor with the
mortar plastic properties. The correlation with compressive strength, economy, and
the ease of use of the device make the pendulum hammer a practical device to aid in
the evaluation of in-place mortar. Drawbacks to the use of the device include the
need for frequent calibration and maintenance.
This abstract accurately represents the content of the candidate's thesis. I recommend
its publication.
Signed
Kevin L. Rens
IV


ACKNOWLEDGEMENTS
Portions of this publication were developed under a grant from the National Park
Service and the National Center for Preservation Technology and Training. Its
contents are solely the responsibility of the authors and do not necessarily represent
the official position or policies of the National Park Service or the National Center for
Preservation Technology and Training.
I am grateful to the following individuals for providing materials, technical advice,
and access to historic structures:
Dr. Margaret Thomson, Chemical Lime Corporation, Henderson, Nevada
Richard Gervais, Dabois Inc., Quebec, Canada
Richard Beardmore, A E Design Associates P.C., Fort Collins, Colorado
James Stratis, Colorado Historical Society, Colorado
Carol Tunner, Historic Preservation Planner, City of Fort Collins, Colorado
Wayne Ruth, Masonry Solutions Inc., Baltimore, Maryland
Glen Boomazian, Integrated Conservation Resources Inc., New York, New York


CONTENTS
Figures
Tables
Chapter
1 Introduction...............................................................1
1.1 Background................................................................1
1.2 A Stress Analysis of a Brick on Two Different Mortars.....................1
1.3 Purpose...................................................................4
1.4 Scope.....................................................................5
2 Literature Review..........................................................6
2.1 Mortar History............................................................6
2.2 The Chemistry of Mortars.................................................7
2.2.1 Chemistry of Lime Mortars...............................................7
2.2.2 Hydraulic and Non-Hydraulic Mortars.....................................8
2.3 Problems with Repointing Historic Masonry................................8
2.3.1 Mortar Stiffness........................................................8
2.3.2 Moisture Migration......................................................9
2.3.3 Crystallization of Soluble Salts.......................................10
2.4 Mortar Testing and Analysis............................................ 10
2.4.1 Relevant American Standardized Tests...................................10
VI


2.4.2 Tests for the Indirect Determination of In-Situ Mortar Strength.........11
2.4.3 Evaluations and Comparisons of In-Situ Mortar Test Methods..............16
2.5 Current Practice in Mortar Evaluation for Historic Preservation Projects.20
2.5.1 On-Site Visual Evaluation..............................................20
2.5.2 Aggregate Gradation, Color, Particle Size and Shape.....................20
2.5.3 Analyzing Mortar Composition in the Laboratory..........................21
2.6 Hardness Testing........................................................22
2.6.1 The Mohs Hardness Scale................................................22
2.6.2 Indentation Tests.................................................... 22
2.6.3 Rebound Tests..........................................................23
2.7 Case Studies............................................................25
2.7.1 Civic Tower of Pavia...................................................25
2.7.2 Conservation of a Historic Stone Building in Trinity College, Dublin....27
3 Materials.................................................................29
3.1 Cement...................................................................29
3.2 Lime....................................................................29
3.3 Sand....................................................................29
3.4 Bricks..................................................................29
3.5 Mortar Formulation and Mixing...........................................30
3.5.1 Mortars for Laboratory Characterization................................30
3.5.2 Mortars for Pier Construction..........................................30
4 Construction of Test Specimens............................................32
vii


4.1 Construction of Piers for Hardness Testing.............................32
4.2 Preparation of Mortar Cubes for Compression Testing....................33
4.3 Preparation of Specimens for Water Vapor Transmission Testing..........34
5 Test Procedures.............................................*...........35
5.1 Sieve Analysis of Sands.................................................35
5.2 Tests for Mortar Plastic Properties....................................35
5.2.1 Mortar Flow...........................................................35
5.2.2 Modified Vicat Cone Penetration.......................................36
5.2.3 Water Retention.......................................................36
5.2.4 Air Content...........................................................38
5.2.5 Water/Binder Ratio....................................................38
5.3 Mortar Hardened Properties.............................................38
5.3.1 Mortar Compressive Strength...........................................38
5.3.2 Water Vapor Transmission..............................................39
5.4 Rebound Number.........................................................39
5.4.1 Preliminary Investigation.............................................40
5.4.2 Experimental Data Collection Procedure................................42
6 Results and Discussion..................................................44
6.1 Sieve Analysis of Sands.................................................44
6.2 Mortar Plastic Properties..............................................45
6.3 Mortar Hardened Properties.............................................46
6.3.1 Mortar Compressive Strength...........................................46
viii


6.3.2 Water Vapor Transmission...........................................48
6.4 Rebound Hardness of Laboratory Specimens............................49
6.4.1 Determination of the Number of Tests Necessary to Characterize Rebound
Hardness..................................................................50
6.4.2 Calibration of the Rebound Hammer..................................52
6.4.3 Rebound Hardness Development over Time.............................53
6.4.4 Hardness of Mortars after 90 Days of Curing........................54
6.4.5 Rebound Hardness versus Compressive Strength.......................56
6.4.6 Rebound Hardness versus Joint Tooling...............................58
6.4.7 Rebound Hardness versus Sand Gradation..............................59
7 Field Tests............................................................60
7.1 Basilica of the Assumption............................................60
7.2 Saint Alphonsus Church................................................61
7.3 Mahan Hall............................................................62
7.4 Chamberlin Observatory............................................. 64
7.5 The Graham Bible House Carriage House..............................67
7.6 B. Jr.s Auto Parts Store............................................68
7.7 The Centennial School Gymnasium Addition.............................69
8 Conclusions and Recommendations........................................71
8.1 Factors Affecting Rebound Hardness....................................71
8.2 Development of Test Methods...........................................71
8.3 Usefulness of Rebound Hardness Results................................72
8.4 Problems and Limitations..............................................72
ix


8.5 Future Work and Recommendations
73
Appendix.................................................................74
Appendix A. Draft Standard Test Method...................................75
Appendix B. Referenced ASTM Standards................................. 79
Appendix C. Ancillary Photographs........................................80
AppendixD. Finite Element Analysis Supporting Materials..................91
Appendix E. Pendulum Hammer Data (abridged).............................105
References..............................................................108
X


FIGURE
1.1 Section chosen for finite element analysis.................................2
1.2 Plot of vertical stresses in a cross section of a brick bearing on two different
mortars...........................................................................3
1.3 Plot of vertical shearing stresses in a cross section of a brick bearing on two
different mortars.................................................................4
2.1 Moisture transmission through cement and lime mortar joints (from Gibbons,
1995).............................................................................9
2.3. The Schmidt type PM pendulum hammer in use.................................12
2.4. Example of the spacing of impact points for determining the rebound value in
accordance with RILEM MS.D.7.....................................................13
2.5. Example of the spacing of impact points for determining the rebound value in
accordance with ASTM C 805...................................................... 14
2.6. The screw pullout test (from BRE Digest 421)...............................15
2.7. Rebound numbers from various mortar mixes in laboratory walls of clay bricks
and calcite bricks (Grieve et al.)...............................................18
2.8. Cross-section of a wall from the Pavia Tower...............................26
5.1. Apparatus Assembly for the Water Retention Test (figure taken from ASTM C
110).............................................................................37
5.2. Rebound readings for a series of successive impacts taken from a test wall at the
University of Colorado at Boulder Structural Engineering Laboratory..............40
5.3. Results of a statistical analysis of preliminary rebound number data (rebound
number as the average of the last 5 of a series of 10 impacts)..................41
5.4. Results of a statistical analysis of preliminary rebound number data (rebound
number as the first impact of a series)..........................................42
6.1. Sand gradation curves.......................................................45
6.2. Compressive strength of mortars over time...................................47
XI


6.3. Comparison of water vapor transmission rates of type O mortars with different
sand types....................................................................49
6.4. Number of samples required when the average of the last 5 of a series of 10
impacts from each location is used............................................51
6.5. Number of samples required when only one rebound is measured from each
location.................................................................. 52
6.6. Rebound number versus time for selected mortars.........................53
6.7. Rebound hardness versus compressive strength for mortars made with the same
sand, lime and cement.........................................................57
6.8. Rebound hardness versus compressive strength for all mortars............58
7.1. Pendulum hammer testing in the undercroft of the Basilica of the Assumption. 61
7.2. South elevation of Mahan Hall, West Point Military Academy..............62
7.3. Poorly bonded pointing mortar at Mahan Hall.............................63
7.4. South elevation of the Chamberlin Observatory............................65
7.5. Southwest elevation of the Graham Bible House Carriage House............67
Xll


TABLE
2.1. Mortar Classification Chart (developed from ASTM C 270, Standard
Specification for Mortar for Unit Masonry)...................................6
2.2. Pointing hardness classification (from RILEM MS.D.7)....................17
2.3. The Mohs Hardness Scale.................................................22
3.1. Mortar Mix Proportions by Volume and Weight.............................30
4.1. Pier specifications......................................................33
6.1. Sieve Analysis Results...................................................44
6.2. Mortar Plastic Properties................................................46
6.3. 90 day and later compressive strength of mortars........................47
6.4. Rate of water vapor transmission of selected mortars....................48
6.5. Development of compressive strength and rebound hardness of mortars from the
first week to the fourth week of curing......................................54
6.6. Rebound hardness of laboratory piers after more than 90 days of curing. All
mortars contain the same cement, lime, and sand..............................55
7.1. Results of petrographic analysis and rebound hammer testing of mortars at
Mahan Hall....................................................................64
7.2. Rebound number test results from the Chamberlin Observatory.............66
7.3. Pendulum hammer test results from B. Jr.s Auto Parts Store..............68
7.4. Hardness results from the three types of wall construction at the Centennial
School Gymnasium Addition.....................................................70


1 Introduction
1.1 Background
Most masonry preservation projects include some degree of pointing work to repair
deteriorated mortars, yet there is no standard methodology for in-place evaluation of
masonry mortars or qualification of repair mortars following repointing. The current
state of practice is to use chemical and petrographic testing to determine mortar
composition. Although these procedures work well, the time and expense required
for chemical and petrographic testing usually limit the investigation to only a few
samples. Many masonry preservation projects have failed where inappropriate
mortars were specified and used for pointing. For example, the use of a stiff
repointing mortar can cause brick spalling and durability problems.
1.2 A Stress Analysis of a Brick on Two Different
Mortars
A finite element model of a brick on a bed of two different mortars was created to
develop an understanding of the stress conditions resulting from the use of overly
stiff pointing mortar. Figure 1.1 shows the section that was modeled.
l


Figure 1.1 Section chosen for finite element analysis.
A cross-section perpendicular to the face of the wall was examined. One brick and
one half of a mortar joint were included. A load was applied to the top of the brick,
to simulate a wall load, and the mortar was restrained from movement in the vertical
direction. The base of this mortar joint is an axis of symmetry, that is, the model is
equivalent to a model with another half of a mortar joint and another brick below this
plane.
The materials were assigned moduli of elasticity (E) to express their relative
stiffnesses. The brick was assigned a value five times that of the soft mortar, and the
hard mortar was given a value 10 times greater than that of the soft mortar.
Six hundred and seventy-two 4-noded quadrilateral elements were used. By default,
a fifth node was added to the center of each element by the program used. A
complete quadratic stress distribution is assumed for each element. These plate finite
elements were assigned a nominal thickness. The computer input and output files for
the analysis, and an explicit account of the assumed stress distribution for the
elements, is included in Appendix D.
2


Figure 1.2 and Figure 1.3 show results of the finite element analysis. Figure 1.2 is a
plot of tensile and compressive stresses in the vertical direction.
Figure 1.2 Plot of vertical stresses in a cross section of a brick bearing on two
different mortars.
The scale on the left associates colors with numerical ranges of stress. The colors
less than zero represent tensile stresses, while the colors greater than zero represent
compressive stresses. Notice that the maximum compressive stress occurs in the
region where the two mortar types meet. The brick rotates toward the softer mortar
about this region, which acts as a fulcrum. The softer mortar deforms downward in
compression, while the harder mortar at the face of the wall deforms slightly upward
in tension. This state of tension at the face of the wall where the hard pointing
mortar and the brick meet could cause the two materials to separate. Such separation
of hard pointing mortar and brick is often noted in the field, inviting water ingress,
freeze thaw damage, and other related problems.
Spalling of brick faces is also common when strong Portland cement mortars are
used to repoint relatively soft masonry. Such brick spalling could be attributable to
3


the tensile stresses at the brick face indicated in Figure 1.2 acting in combination
with shearing stress. Figure 1.3 shows the shearing stresses in the vertical direction
that were determined using the same model as above.
Figure 1.3 Plot of vertical shearing stresses in a cross section of a brick bearing on
two different mortars.
1.3 Purpose
The objective of this research program was to develop a method for the use of the
Schmidt Type PM pendulum hammer to evaluate and characterize in-place masonry
mortars. The method is intended to be used as an aid in the selection of pointing
mortar, for quality control of new construction, and for general nondestructive
evaluation of in-place mortar. In particular, it is hoped that the test method
developed here will be used to choose appropriate pointing mortars for historic
structures, thus avoiding problems such as the stress concentrations that are
described in section 1.2.
4


1.4 Scope
Mortar properties were investigated and compared to rebound hardness data. These
properties include mortar compressive strength, masonry unit compressive strength,
water vapor transmission, and mortar plastic properties. Eight different mortar
formulations were investigated, including seven mortars containing different ratios
of lime to cement as the binder, and one mortar using hydraulic lime as the binder.
Pendulum hammer rebound data collected from laboratory test piers was analyzed in
order to develop a test methodology that provides statistically significant results.
The rebound data was then compared to the mortar properties to determine whether
or not the rebound data could be used as an indicator of those mortar properties. A
set of field trials were conducted on historic structures in Colorado and across the
United States to compare the results of pendulum hammer testing with other
evaluation methods and to develop guidelines for field use of the device. The
experience gained from the field testing and the results of the laboratory comparisons
were used to develop a test method for hardness determination with the pendulum
hammer. This method includes guidelines for use of the equipment and provides a
method for obtaining statistically significant results. A standard test method for this
technique has been drafted and is included in Appendix A.
5


2 Literature Review
2.1 Mortar History
Masonry mortars of lime, sand, and water were used as early as 2450 BC (Speweik,
1997). The slaking of lime was well known to Roman culture by the time of
Vitruvius (20 AD). Vitruvius gives mortar recipes in his "Ten Books of
Architecture" that describe the state of the art in mortars until the introduction of
Portland cement in the 1870's (Sickels-Taves, 1997). The formulation of mortar with
lime and sand was most likely passed from Greece to Italy, where the practice was
refined and used in brick and tile construction (Blake, 1947). The Romans had easy
access to natural deposits of pozzolana, a volcanic product containing silica and
alumina, which further strengthened their mortars (Blake, 1947).
Since about 1870, mortars have been made from various combinations of lime,
Portland cement, and sand. Buildings constructed in the United States from 1870 to
1930 contain mortars of various ratios of lime and cement, but after 1930 one part
lime putty to one part Portland cement was the most common mix (Speweik, 1997).
In 1954 a mortar classification scheme was introduced to distinguish between
mortars of different mix proportions and strength. The letters M, S, N, O, K (from
MaSoNwOrK), distinguish the five mortar types.
Table 2.1. Mortar Classification Chart (developed from ASTM C 270, Standard
Specification for Mortar for Unit Masonry).
Mortar Type Proportion by volume Compressive Strength Minimum Typical*
Cement Lime Sand psi psi
M 1 1/4 2A to 3 times 2500 6400
S 1 1/2 the sum of the volumes of 1800 3625
N 1 1 cement and 750 1850
O 1 2 lime 350 540
K 1 3 75 190
*Typical strengths are from Frey, 1975, for mortars prepared and tested in the
laboratory per ASTM C 270.
6


2.2 The Chemistry of Mortars
2.2.1 Chemistry of Lime Mortars
Traditional lime mortars are manufactured by burning limestone or other material
containing calcium carbonate to produce quicklime, and then slaking the quicklime
to produce lime putty or dry hydrated lime. Quicklime is made by heating calcium
carbonate to over 850 degrees Celsius. At this temperature, the heat drives off carbon
dioxide from calcium carbonate leaving calcium oxide:
CaC03 + Heat (850 C)^ CaO + C02 (gas)
(Calcium Carbonate) (Calcium Oxide) (Carbon Dioxide)
The calcium oxide, or quicklime, is then slaked with water. Slaking is an exothermic
reaction of water and quicklime that produces calcium hydroxide, referred to as lime
putty or dry hydrated lime depending on whether the end product is dry or left in
excess water.
CaO + H20 Ca(OH)2 + Heat
(Calcium Oxide) (Calcium Hydroxide)
Lime putty is traditionally sieved and left to mature in a pit or box for 2 weeks to
one year. During this maturation process, the slaking reaction continues and the
particles of the putty will decrease in size, resulting in a higher quality material
(Gibbons, 1995). The lime putty is then mixed with sand at the site prior to use. The
mortar then hardens as it reacts with carbon dioxide from the atmosphere to reform
the original limestone component, calcium carbonate. This carbonation reaction may
take several years to reach completion.
Ca(OH)2 + C02 CaC03 + H20
(Calcium Hydroxide) (Carbon Dioxide) (Calcium Carbonate)
7


2.2.2 Hydraulic and Non-Hydraulic Mortars
The reactions discussed in section 2.2.1 describe non-hydraulic mortar, as
distinguished from mortar containing hydraulic cements. Hydraulic cements harden
by incorporating water into the molecular structure of the cement. Portland cement,
derived from firing alumina silicate (clay) with calcium carbonate (limestone), is an
example of an hydraulic cement. Roman pozzolana, similar to Portland cement in
that it contains alumina and silica, also provides hydraulic qualities to mortar (Blake,
1947). Lime can also have hydraulic properties when it is made from argillaceous,
or clayey, limestone. These hydraulic limes are further classified as feebly,
moderately, and eminently hydraulic limes (Gibbons, 1995). Hydraulic mortars have
the advantage of faster hardening and the ability to set under water.
2.3 Problems with Repointing Historic Masonry
Historic masonry structures require periodic pointing as mortars decay. In older
structures built with soft mortar, the mortar is intended to be a sacrificial material,
relieving brick stresses by absorbing movement and protecting the bricks from water
by providing a path for the movement of moisture (Speweik, 1997). A mortar that
plays this sacrificial role for the bricks of a building will wear out in time, typically
in about fifty years. The exposed face of the mortar will recede, especially when the
mortar is exposed to frequent wetting and drying cycles (Williams, 1979).
The repointing of masonry consists of removing the mortar in a wall to some depth,
typically the outer three-quarters of an inch to one and a half inches, and replacing it
with fresh mortar. Historic preservationists strive to match the color, ingredients,
and material properties of the original mortar when repointing historic structures
(Williams, 1979; Sickels-Taves, 1997).
2.3.1 Mortar Stiffness
Since the introduction of Portland cement, many structures have been damaged by
the use of repointing mortars that are stronger and more dense than the original
mortar. Flexible lime mortars of low elastic modulus allow for small movements in
walls from settlement, temperature change, etc., whereas stiff, relatively inflexible
mortars with high cement content force the masonry units to absorb this movement
(Sickels-Taves, 1997). This rigid restraint can cause bricks and stones to crack and
spall, particularly in older buildings with bricks of low strength (Williams, 1979).
8


This effect is exacerbated in repointed masonry because the replacement of the
original mortar is partial. Thus the bricks are supported by two materials of different
strengths, resulting in stress concentrations as described in section 1.2.
2.3.2 Moisture Migration
A further problem that can be caused by repointing with cement mortar is a change
in the migration paths of water through masonry. Lime mortars are porous and allow
moisture ingress and egress, while cement mortars are much less porous (Williams,
1979; Speweik, 1997). Non-porous mortars can force moisture transmission through
bricks and thus cause the bricks to deteriorate rapidly (Gibbons, 1995). Figure 2.1
illustrates the migration of moisture in cement and lime mortar joints.
moisture evaporates through stone
. and facte of stone weathers back
moisture evaporates through marfer
and face of joint weathers back
Figure 2.1 Moisture transmission through cement and lime mortar joints (from
Gibbons, 1995).
Moisture transmission through bricks is also encouraged by cement mortars because
cement mortars can shrink upon setting, and can thus leave spaces for water to
penetrate a wall at the mortar-brick interface (Speweik, 1997).
9


2.3.3 Crystallization of Soluble Salts
Mortars can also be damaged by the presence of soluble salts (Gibbons, 1995; Duffy
et al., 1993; Speweik, 1997). Cement mortars have higher levels of soluble salts
naturally present in their components than do lime mortars (Duffy et al., 1993).
Transfer of these salts to adjacent porous stones or bricks can result in damage as the
salts crystallize (Gibbons, 1995). Salts from natural sources and pollution can also
darhage masonry. In coastal areas, windbome salts can contribute to the dissolution
of calcium compounds. Crystallizing salts in the pores of masonry develop
destructive pressures (van der Klugt, 1991).
2.4; Mortar Testing and Analysis
I
2.4.1 Relevant American Standardized Tests
i
All American Society for Testing and Materials [ASTM] methods that are
mentioned in this work are listed in Appendix B. There are standardized tests
commonly used for the analysis of mortars prior to and during construction ASTM C
780, Test Method for Preconstruction and Construction Evaluation of Mortars for
Plain and Reinforced Unit Masonry), but there are no accepted American standards
for the analysis of in-situ mortar. ASTM subcommittee E6.24 on Building
Preservation and Rehabilitation Technology was formed in the 1980's to "develop
standards in the technology of conservation, preservation and rehabilitation of
buildings and structures (Kelly and Slaton, 1994)." The United States Secretary of
the Interior's Standards for Rehabilitation and Guidelines for Rehabilitating Historic
Buildings recommends that "old mortars be replaced with mixtures that match the
original in strength, composition, color, and texture (Doebley and Spitzer, 1996)."
There are, however, tests that are related to in-situ mortar strength. The bond wrench
test (ASTM C 1072, Test Method for Measurement of Masonry Flexural Bond
Strength) uses a special wrench to test the bond strength of one masonry bed joint.
This test has been modified for field use using a special clamping head.
The in-place push test (Uniform Building Code [UBC] Standard Number 21-6) is
used to test the shear resistance of one masonry unit by removing a brick and
inserting a hydraulic piston. The head joint next to an adjacent brick is removed and
a load vs. deflection test is performed. Figure 2.2 shows a schematic of this test.
This test can also be performed by removing the head joints on either side of a brick
10


and using a small flatjack in one of the empty joints to apply the load (Schuller and
Suprenant, 1994).
There are tests available for the indirect in-situ determination of mortar compressive
strength, but none of these are standardized in the United States. A discussion of
these tests follows.
2.4.2 Tests for the Indirect Determination oiln-Situ
Mortar Strength
In-situ tests that can be correlated to the strength of mortar include the pendulum
hammer rebound test, the pull-out test, penetration tests, and drilling resistance tests.
The results of these tests are correlated empirically to strength.
Figure 2.2. Schematic of the in-place push test (from Suprenant and Schuller, 1994).
The pendulum hammer rebound test
The pendulum hammer rebound test uses a Schmidt type PM hammer to obtain a
measure of the hardness of mortar. Figure 2.3 shows the pendulum hammer in use.
11


Figure 2.3. The Schmidt type PM pendulum hammer in use.
A test method using the Schmidt hammer to evaluate mortar quality has been
adopted as a suggested method by the International Union of Testing and Research
Laboratories for Materials and Structures (acronym: RILEM), RILEM MS.D.7:
Determination of Pointing Hardness by Pendulum Hammer. ASTM C 805, Test
Method for Rebound Number of Hardened Concrete, offers a standard method for
the evaluation of concrete using the higher energy spring loaded Schmidt hammers.
The International Society for Rock Mechanics (ISRM) has adopted a method for
using the Schmidt hammer to measure rock hardness, the Suggested Method for
Determination of the Schmidt Rebound Hardness.
The RILEM method specifies that nine rebound readings are to be taken at different
locations with the type PM pendulum hammer. Specifically, RILEM MS.D.7 states:
"Perform nine measurements uniformly divided over the area to be judged, e.g. three
12


times three measurements horizontally and vertically at one third from the
boundaries of the area or the edges of the panel to be measured." The median of
these readings is taken as the rebound value. Figure 2.4 shows a graphical
interpretation of this method.
L/4 L/4 L/4 L/4
Figure 2.4. Example of the spacing of impact points for determining the rebound
value in accordance with RILEM MS.D.7.
ASTM C 805 specifies that 10 individual hits are to be made in a test area, and the
average of these (less the outliers) is taken as the rebound number. The test area is to
be at least six inches in diameter, and the concrete is to be at least four inches thick
and fixed within a structure. Figure 2.5 shows an example of the spacing of impact
points for ASTM C 805. This method is intended to identify regions of poor quality
or deteriorated concrete in a structure by testing several areas (ASTM C 805).
13






The test area is to be at
least 6 inches in diameter

Impact points no two impact points are to be closer than one
t inch to each other
Figure 2.5. Example of the spacing of impact points for determining the rebound
value in accordance with ASTM C 805.
The ISRM method is to impact a rock sample at least 20 times. The impact locations
are to be separated by at least the diameter of the plunger. The test surface is to be
smooth and flat, and away from any local discontinuity or cracks in the rock. Small
specimens are to be securely clamped to a rigid base. It is noted that the hammer
should be calibrated using a calibration test anvil from the manufacturer. A
correction factor is calculated as the specified standard value of the anvil divided by
the average of ten readings on the calibration anvil. The rebound value of the rock is
taken as the average of the upper 50% of the values obtained, multiplied by the
correction factor (International Society of Rock Mechanics, 1978).
14


Pull-out tests
Pull-out tests involve measuring the load required to pull an anchor bolt or a helical
tie out of a mortar joint. Figure 2.6 shows a schematic of a pull-out test.
Figure 2.6. The screw pullout test (from BRE Digest 421).
This test is also known as the helix test. The Building Research Establishment
(BRE) of Britain has published a digest describing this screw pull out test, including
the technical background, calibration, and interpretation of results (BRE Digest 421).
Pilot holes at least thirty millimeters deep are drilled into the mortar at ten locations
within a two meter square area. The holes are to be of the size to make an
interference fit with the ties that have been selected. The helical ties are then tapped
into the holes with a hammer. A proof-loading device is attached to the tie, and the
tie is loaded to failure. The peak load recorded during the test is the pullout load.
The pullout load can be related to the mortar strength using a curve that correlates
pullout loads to mortar cube strength, or used as means to compare the mortar of
various regions of a structure. There are other pull-out tests for other types of
masonry connectors, such as anchors and expansion bolts.
Penetration tests
Two types of penetration tests are common. The Windsor Penetrometer uses an
explosive charge to propel a probe into a mortar joint (Grieve et al.). The probe
penetration is measured and correlated to mortar cube compressive strength. Two
probes are available for different ranges of compressive strength. A silver-colored
probe of 3/16 inch diameter is used for materials having a compressive strength in
15


the range of 3000 to 5000 pounds per square inch. A gold-colored probe of 5/16
inch diameter is used for materials of lower strength.
The Pin Penetration Resistance test (PPR) uses a spring-loaded rod to drive a steel
pin into the test material. The pin diameter is 0.140 inches and it projects 0.3 inches
beyond its holding rod. The pin strikes the test material and leaves an indentation.
A depth gauge is used to measure the depth of penetration, which can be correlated
to mortar compressive strength.
Recently, Italian researchers have developed a penetration test that uses the
controlled impact energy of a Schmidt rebound hammer to drive a steel pin into
mortar (Felicetti and Gattesco, 1998). A series of blows are used to drive the pin
into the joint and the penetration depth is measured with each successive blow. This
test is similar to the Standard Penetration Test that is commonly used for in-situ soil
testing in the United States. Felicetti and Gattesco conducted an extensive laboratory
study correlating the results of this test to the compressive strength of mortars. This
test has the advantage of measuring the resistance of the mortar at many discrete
depths, but does leave a hole in the mortar.
Drilling resistance tests
The drilling resistance test correlates the energy required to drill a hole of a certain
depth to the strength of the mortar. One technique using this principle is the PNT-G
penetrometric method (de Vekey and Sassu, 1997). This method uses a drill and an
energy gauge manufactured and calibrated for this purpose.
Another drilling resistance test has been developed in Czechoslovakia by Vaclav
Kucera. This test uses a hand turned impact drill that is held against the test material
with a constant, measured force. The strength is taken as a function of the force, the
number of turns, and the depth and diameter of the hole. A calibrating table is made
from tests of mortars of known strength.
2.4.3 Evaluations and Comparisons of In-Situ
Mortar Test Methods
L. J.A.R van der Klugt of the Dutch organization TNO Building and Construction
Research has adapted the Schmidt pendulum hammer to test the hardness of pointing
mortar (van der Klugt, 1991).
16


The lightest Schmidt pendulum hammer was fitted with a 5 mm diameter shaft and
impact head for mortar testing. The instrument was equipped with an adjustable
distance holder and a spirit level vial to keep it vertical when testing the slanting
masonry of a windmill. Many readings (usually nine) were taken on all test
specimens, and rather than the average, the median of the readings was taken as the
rebound number.
Tests were conducted on walls built in a laboratory with mortars of different
formulations. Fourteen mortars were tested, ranging from 1:3 Portland cement to
sand to 1:3 lime to sand. The hardness of these mortars was measured weekly as the
mortar cured.
Tests were conducted to compare the hardness of mortars that were dry to those
which were saturated. The effect of three different brick types on the hardness
number of the mortars was also investigated. Some qualitative results of interest
include:
Bed joints were stronger than head joints.
Mortar hardness in walls built with molded bricks and those built with extruded
bricks were equivalent. The same mortar in walls built with sand-lime bricks
gave significantly lower rebound numbers.
Mortar in walls which were kept damp during curing gave higher rebound
numbers than mortar in walls which were kept dry.
Mortar in walls which were measured when wet gave lower rebound numbers
than when measured dry.
A classification table of rebound number as a function of mortar quality is offered.
This table is shown in Table 2.2.
Table 2.2. Pointing hardness classification (from RILEM MS.D.7).
class hardness indicated quality
0 (zero) <15 very soft
A 15-25 soft
B 25-35 moderate
C 35-45 normal
D 45-55 hard
E 55 very hard
Grieve, Marshall and Willoughby conducted a study comparing three types of mortar
strength tests for the Ontario New Home Warranty Program (Grieve et al.). The
Windsor Penetrometer test, the pendulum hammer rebound test, and the pin
17


penetration resistance test were evaluated and compared. Each technique was used
to test five different mortar mixes in laboratory test walls. The object was to find a
simple, in-situ, non-destructive method for the evaluation of mortar.
The Windsor Penetrometer was found to de difficult to use, unreliable, and
dangerous for mortar testing. The Schmidt hammer was found to be.convenient to
use and to give reasonable results. However, the rebound numbers were found to be
more affected by brick type than mortar mix proportions. Figure 2.7 shows the
rebound numbers recorded for test walls of clay and calcite bricks. Notice that the
values for the calcite bricks are consistently lower yet show the same trend for the
different mortar mixes.
Rebound Number versus Mix
From Grieve et al. For Clay and Calcite Bricks
-clay
. calcite1
Mortar Mix: Parts sand to one part cement
Figure 2.7. Rebound numbers from various mortar mixes in laboratory walls of clay
bricks and calcite bricks (Grieve et al.).
The results were in agreement with the quality scale suggested by the RILEM
standard MS.D.7: Determination of Pointing Hardness by Pendulum Hammer.
Preparation of the mortar surface with a surface grinder to provide a flat striking
surface provided more consistent results. The authors note that the main drawback
to this test is that the striking head of the hammer is of a fixed length. The authors
suggest that different impact heads would be useful. Another drawback was the
18


problem of accurately striking the surface of a curved joint. Despite this drawback,
the vast majority of mortar joints are accessible to this instrument and it should prove
eminently suitable for site quality control and laboratory research (Grieve et al.).
The Pin Penetration Resistance test was found to be simple to operate but required
meticulous care in measuring the gap between the device and the mortar joint.
Testing with this device was much slower than with the Schmidt Hammer. The
results obtained with the PPR meter were similar to those of the Schmidt hammer in
accuracy, and in that the test was more sensitive to brick type and joint tooling than
to mortar mix proportions.
The Schmidt hammer and the PPR meter yielded good results on the test panels. The
Schmidt hammer gave reasonable results but has the problem of not being able to
test at various depths. The type PM hammer and the PPR meter test demonstrate that
the absorption characteristics of bricks play a dominant role in the in-situ
measurement of the mortar. Grieve et al. suggest that the calcite bricks did not
adequately absorb the water in the mortar mix, resulting in significantly lower joint
strengths.
De Vekey and Sassu compared the PNT-G drilling resistance method and the Helix
screw pullout methods for the determination of the compressive strength of masonry
mortars (De Vekey and Sassu, 1997). The researchers statistically correlated the
results of these two methods to results of mortar cube compression tests. The PNT-
G penetrometer was found to give good results for weak and very weak mortars.
The Helix pull-out test gave the best results for medium and strong mortars. Neither
method yielded good results for very strong mortars. De Vekey and Sassu conclude
that both of these tests give more reliable and accurate values for mortar strength
than other techniques such as the Windsor Penetrometer, the Schmidt hammer, and
strength estimates based on chemical analysis.
RILEM committee 127 MS conducted round robin tests to verify and calibrate
proposed test methods, one of which was the rebound hammer test to measure the
surface hardness of porous materials (Binda, 1997). Binda reports that the hammer
rebound hardness measurement can be correlated to strength for homogenous,
isotropic, elastic materials. Although this does not describe masonry, the rebound
number can be correlated to the condition of the material. To calibrate the method to
different masonry types and to understand if extended applications could be made,
three series of measurements were carried out in Italy on mortar, brick, and stone
laboratory specimens in Milan. Results show that the technique can be used to
classify repointing mortars of cement, cement-lime, or hydraulic lime mortars. The
tool is not able to classify lime putty mortars as they are too soft. A pendulum
19


hammer that delivers less impact energy should be produced for these mortars
(Binda, 1997).

2.5 Current Practice in Mortar Evaluation for
Historic Preservation Projects
2.5.1 On-Site Visual Evaluation
There are some simple rules of thumb that are applied frequently to evaluate in-place
mortar. Lime mortars appear white and often will leave white dust on one's finger
when rubbed. Cement mortars have a grey color and should not produce dust when
rubbed. Further clues to the age of the mortar can be gleaned from the nature of the
bricks (Sickels-Taves, 1997). Non-standard sizes indicate old mortars (pre 1899), as
do porous, handmade bricks. Later, twentieth century bricks have been produced
with manufacturing techniques that have eliminated such variability. It is common
to scrape the mortar with a hard object such as a key to gain a rough idea of the
hardness of the mortar (Speweik, 1997).
2.5.2 Aggregate Gradation, Color, Particle Size and
Shape
The gradation of the original aggregate of a mortar can be determined by acid
digestion of the binder matrix followed by sieving with ASTM standard sieves
(Speweik, 1997). Care must be taken to account for the possibility that limestone
aggregate is present, as it can be dissolved along with the binder. The sand used for
new pointing mortar should have a similar gradation curve. The shape of the sand
particles also is important to the behavior of the resultant mortar. Round particles, as
found in natural river or beach sand, provide less cohesion than sharp, angular
particles such as those found in pit sand or manufactured sand. However, natural
sands produce more workable mortars that may more closely approximate the
behavior of an historic mortar (Speweik, 1997).
The shape of sand particles can be approximated by examination with a hand lens
(Sickels-Taves, 1997). If an in-depth study of the source of a sand is desired, more
advanced methods are available, such as optical stereology, petrography, and
mineralogical analysis of the aggregate (Sickels-Taves et al, 1997). Also, the U.S.
20


Heritage Group, Inc., has established the "National Sand Library" in Chicago,
Illinois, where sands from across the country are catalogued geographically for the
purpose of providing matching sands for mortar.
It is vital to the historic preservationist to match the color of pointing mortar to the
original mortar (Higgins, 1984). The residue of Portland cement is medium to dark
gray; clay produces a reddish to light tan color; and the residues of natural, non-
manufactured cement are brown (Sickels-Taves, 1997). Additives such as lumps of
partially burned lime, animal hair, cows blood, egg whites, brick dust, seashells, or
carbon from inadvertent mixture during the firing of the quicklime may be present
(Speweik, 1997). The color of the replacement mortar should match the
unweathered original mortar. This can usually be achieved by closely matching the
original aggregate and binder, but color additives are available (Higgins, 1984).
2.5.3 Analyzing Mortar Composition in the
Laboratory
Treatment of a mortar sample with hydrochloric acid can indicate the presence of
lime or cement binders. Vigorous bubbling and amber color indicate the presence of
lime, while weak agitation and a murky green color indicate cement (Sickels-Taves,
1997).
By means of analytical chemistry and petrography, hardened mortar can be studied
to reveal (Erlin and Hime, 1985):
Composition and proportions
Admixtures
Air-void systems
Hydration of the Portland cement, hydration reactions, and water/cement ratio
Textural features related to original mortar consistency and workmanship
Water migration paths
Chemical and physical soundness of the components
Features of the mortar related to performance
More advanced chemical techniques can be used in tandem with petrographic studies
to quantitatively evaluate mortar samples. ASTM C 1324, Standard Test Method for
Examination and Analysis of Hardened Masonry Mortar, provides explicit
guidelines for chemical and petrographic analysis.
21


2.6 Hardness Testing
Tests commonly used to measure the hardness of rocks, concrete, and similar
materials include scratch tests, indentation tests, and rebound tests. Hardness is
considered as a material behavior rather than a fundamental physical property of a
material. Hardness is related to or is a function of resistance to scratching, Youngs
modulus, brittleness, etc. (Atkinson, 1973).
Hardness values are expressed in empirical units relevant to the particular test used.
2.6.1 The Mohs Hardness Scale
The Mohs hardness test gives a comparative value of hardness by specifying a
hardness scale from one to ten. Each hardness value is represented by a mineral,
with diamonds assigned the value ten and talc assigned a value of one. The Mohs
scale is given in Table 2.3. The minerals with higher hardness numbers are capable
of scratching those with lesser numbers (Atkinson, 1973).
Table 2.3. The Mohs Hardness Scale.
Mohs Hardness Scale
Mohs Number Mineral
1 Talc
2 Gypsum
3 Calcite
4 Flourspar
5 Apatite
6 Feldspar
7 Quartz
8 Topaz
9 Corundum
10 Diamond
2.6.2 Indentation Tests
Indentation methods are common for metals. The Brinell test uses a spherical metal
ball applied to the surface with a specified force. The hardness value is a function of
22


the geometry of the indented surface and the applied load (Boyer, 1977). An
adaptation of this type of slow-load indentation device was developed for testing
rocks (van der Blis, 1970). Empirical relations were found for the Brinell hardness
number and the elastic moduli of the rock. Three methods employing this principle
for concrete testing are the Williams Testing Pistol, the Frank Spring Hammer, and
the Einbeck Pendulum Hammer. These methods can be used to predict strength with
an accuracy of twenty to thirty percent (Malhotra, 1976). The Windsor Penetrometer
and the Pin Penetration Resistance Meter described previously in Section 2.4.2 are
examples of indentation testers suitable for mortars.
2.6.3 Rebound Tests
The Shore Hardness Test
In 1911, Shore described a scleroscope method for determining the hardness of
metals. This method involves measuring the height of rebound of a hardened steel
hammer that is dropped on the test metal (Malhotra, 1976). In current practice, the
Shore Scleroscope hardness test consists of dropping a diamond-tipped hammer from
a fixed height and measuring the rebound. The hammer, weighing about two grams,
slides inside of a glass tube (Boyer, 1987). The Shore Hardness test is also used for
comparing the hardness of rocks using a diamond or tungsten carbide hammer. The
mean of 20 readings is usually taken as the Shore hardness of the rock (Rabia and
Brook, 1979). Rabia and Brook found that the Shore Hardness values were a
function of the volume of the specimen. The hardness values increased as the
volume of the specimen was increased to a value of about 40 cubic centimeters, and
then remained constant as volume was further increased. Porous rocks were found to
be harder when dry than when wet (Rabia and Brook, 1979).
The Schmidt Hammer
In 1948 Ernst Schmidt developed the Schmidt hammer for measuring the hardness of
concrete by the rebound principle (Malhotra, 1976). The hammer uses a spring to
drive a steel plunger against the specimen with a specified energy. The maximum
rebound of the plunger is measured on a scale attached to the frame of the device
(ASTM C 805).
Malhotra presents a method for calibrating the Schmidt hammer to specific concrete
mixes. Cylinders are cast with the mix of interest and tested with the hammer over
time as they cure. The cylinders are restrained in a compression tester at fifteen
percent of their yield strength, and tested at several locations. The cylinders are then
23


crushed in the usual manner and the rebound numbers are plotted versus compressive
strength. The accuracy of the estimated compressive strength using this technique in
controlled laboratory conditions is ten to fifteen percent, and twenty-five percent in
the field (Malhotra, 1976). Malhotra notes that that the accuracy of field testing
with the Schmidt hammer is affected by the following factors:
1. Smoothness of surface under test.
Troweled surfaces give scattered results. The Swiss Federal Materials Testing and
Experimental Institute recommends that the Schmidt hammer be used only on
concrete which has been cast against forms.
2. Size, shape, and rigidity of the specimen.
Less rigid specimens will give a lower rebound number.
3. Age of test specimen.
The rebound number increases for the first seven days, and then remains relatively
constant, despite increases in actual compressive strength.
4. Surface and internal moisture condition of the concrete.
Surface-dry concrete gives higher rebound numbers than surface-wet concrete.
5. Type of coarse aggregate.
For equal compressive strengths, concretes with limestone aggregate test lower than
concretes with gravel aggregate.
6. Type of cement.
High-alumina cement concrete can give strengths 100% higher than those obtained
using a calibration chart based on concrete made with ordinary Portland cement.
7. Carbonation of concrete surface.
Surface carbonation significantly increases the rebound number.
These variations demonstrate the difficulty involved in comparing the results of the
Schmidt hammer test from one location with those from another location.
24


Poole and Farmer (1980) conducted a study to examine the consistency and
repeatability of the Schmidt hammer test on exposed rocks and to determine the best
procedure for recording rebound values. For this statistical study, four hundred
rebound values were taken with the Schmidt hammer on a large block of dolerite in
the laboratory, and similar data sets were taken at field locations. The data consisted
of ten or fifteen impacts at one location. One set of field data was taken on a
continuous scan line at oneimeter intervals along a 280 m length of tunnel in
calcareous, sandy limestone. The other set of field data was taken on a 200 mm
squared grid laid out on a limestone abutment. The researchers concluded that the
recorded rebound values were statistically representative of the rock only for the
rebound values taken after the third impact in a series. The variation was excessively
high for the initial three impacts at any point. The F-test was used to determine
statistically if the sample sets were equivalent. The data from one of four rocks
tested exceeded the F-test statistic. It is recommended that to obtain a repeatable
rebound value, the peak value from at least 5 successive rebounds should be selected.
2.7 Case Studies
2.7.1 Civic Tower of Pavia
The collapse of the Civic Tower of Pavia in 1989 prompted the researchers to study
the degradation of the mortar of this medieval construction (Baronio and Binda,
1991). It was suspected that mortar deterioration due to pollution had caused the
collapse. Construction of the tower began in the eleventh century and a belfry of
granite and brick masonry was added in the sixteenth century.
Samples were taken from the ruins of the tower and chemical, petrographic, and
mechanical analyses of the material were performed. The mortars present in the
structure were from many different construction phases. All were lime mortars with
lime to sand ratios ranging from 1:3 to 1:5. They appeared to be well-carbonated
lime putty mortars. The medieval rubble walls were 2.8 meters thick on the ground
floor, and the cladding thickness was an average of 0.15 meters thick. Figure 2.8
shows a cross section of the wall.
25


1
Figure 2.8. Cross-section of a wall from the Pavia Tower.
The deterioration of the mortars was localized, mainly in the cladding of the
structure. Baronio and Binda conclude that mortar degradation was not the cause of
the collapse. Degradation was evident only in the cladding material of the structure,
while the mortars at depth were strong and undamaged.
A procedure is outlined for the testing of mortar samples taken from existing
structures. The proposed testing procedure is as follows:
1. Map the various mortar types using all geometrical, stratigraphical, and historic
information available.
2. Mortar should be sampled at the surface and at depth. The sampling should
reflect the nature of the investigation.
3. Chemical analyses can be used to detect the nature of binder and aggregates, the
degree of carbonation of the lime, and the presence of different soluble salts and
sulfates.
4. Microscopic investigation of polished specimens will give a sure indication of
the hydraulic nature of the binder.
5. Thermal treatment of siliceous aggregates allows the determination of the
aggregates grain size distribution.
26


6. Petrographical and mineralogical analyses are used to confirm the chemical
analyses and can be useful to find the source of the original materials.
7. Examination with a microscope can be used to determine the grain size
distribution when the aggregate is of a calcareous or mixed nature.
8. Mechanical tests, when possible, should be used to determine the material
properties of the mortar.
This research was intended to be useful to the future establishment of rules and codes
for the testing of materials sampled from existing structures. The results obtained
show that it is possible to thoroughly characterize historic mortars using a well
defined investigative procedure.
2.7.2 Conservation of a Historic Stone Building in
Trinity College, Dublin
A conservation project was undertaken to restore the Regent House of Trinity
College (Duffy et al., 1993). The stonework of the Regent House facade had been
badly damaged by atmospheric pollution. In 1989, Trinity college began to clean the
blackened and decayed stonework. It was decided that the entire building was to be
repointed.
Previously, the building had been repointed with black colored mortar to match the
blackened stones. To aid in mortar selection, research into the chemical and
mechanical properties of various mortars was conducted. The goal of the testing
program was to select a mortar that would fulfill the necessary functions and not
damage the stone.
The researchers wanted to use a mortar that would prevent the ingress of moisture
and pollutants yet allow the building to dry properly, accommodate movement, and
not contain harmful chemicals that would further deteriorate the stone. Impermeable
mortars cause water to move through the stones, exposing them to frost and chemical
damage. Soluble calcium, sodium, and magnesium salts present in stone and mortar
will change their hydrated form and thus their volume. This can lead to cracks that
allow further ingress of salt-rich water and more crystals. Another problem to avoid
is the loss of bond between mortar and stone due to shrinkage of a wet mortar mix
and overly strong mortars.
The authors conducted a testing program on a series of different mortar mixes to
determine which mortar would best meet their needs. The mortars selected for
testing were various mixes of well aged slaked lime putty and cement, a mortar
27


containing PFA, a pozzolanic material, in place of cement, a mortar containing
barium sulfate in place of lime, a sand and cement mortar with plasticizer, and a
plain sand and cement mortar.
The mortars were tested for compressive and flexural strength, and chemical tests
were performed to determine the levels of soluble salts. The mechanical tests
revealed that the simple sand and cement mortar was the strongest, followed by the
lime and cement mortar, and the cement and plasticizer mortar. The barium mix was
the weakest. The cement plasticizer mix had the least electrical conductivity and
least amount of soluble salts present after 28 days, while the cement and lime mix
was the most conductive and contained the most soluble salts after 28 days. The
cement and plasticizer mix was chosen for the repointing for its appropriate
weakness and chemical properties.
Summary
The stress analysis of a brick on a bed of two mortars, the discussion of moisture
migration, and review of problems associated with inappropriate pointing mortar
were presented to demonstrate a need for in situ mortar analysis. The discussion of
mortar test methods and the review of the literature were presented to provide
context for the research presented here. With this background established, the
methods, data, and results of the research program undertaken to develop the
pendulum hammer test method will be presented.
28


3 Materials
3.1 Cement
The same Portland cement was used for the seven mortar formulations containing
cement binder. Portland cement meeting the requirements of ASTM C 150,
Standard Specification for Portland Cement for Type I cement was used.
3.2 Lime
Three types of lime were used.
1. Type S hydrated lime meeting the requirements of ASTM C 207, Standard
Specification for Hydrated Lime for Masonry Purposes.
2. Lime putty acquired from the Chemical Lime Company of Henderson, Nevada.
Hydraulic Lime acquired from France (Blanche Hydraulique Naturelle Pure, from
NHL pure, France).
3.3 Sand
Three types of sand were used.
1. Quikcrete medium sand. This sand will be referred to as fine sand.
2. Quikcrete all-purpose sand. This sand will be referred to as play sand.
3. A mixture of Quikcrete all-purpose sand and #8 aggregate. This sand will be
referred to as coarse sand.
3.4 Bricks
Two types of bricks were used to construct the laboratory test piers.
1. Reclaimed Boulder Bricks. These bricks were reclaimed from turn of the
century structures in Colorado. These hydraulically pressed, molded bricks were
made by a local manufacturer.
2. Yellow Clay Bricks. These bricks are typical modem extruded and cored bricks.
29


3.5 Mortar Formulation and Mixing
Two types of mortar mix designs were used and are described in the following
section.
3.5.1 Mortars for Laboratory Characterization
The mortars used for the laboratory characterization of plastic and hardened
properties were mixed according to ASTM C 270, Standard Specification for Mortar
for Unit Masonry. The only exception to this specification was that larger batch
sizes were prepared. For this test program, each batch was proportioned using a sand
weight of 7200 grams rather than the 1,400 grams required by ASTM C 270. Mortar
was prepared in a benchtop mixer with a vertical paddle. The mortar types S, N, O,
and K were proportioned to conform to ASTM C 270. The hydraulic lime mortar
was proportioned according to the recommendations of the distributor. These
recommendations conform to historic formulations for mortars with no Portland
cement. Table 3.1 shows the mix proportions by volume and weight of the eight
mortar types tested.
Table3.1. Mortar Mix Pro portions by Volume and Wei ght.
Mix Mortar Type Mix Proportions by Volume Mix Proportions by Weight (grams)
Cement Lime Sand Cement Lime Sand
A Type S 1 1/2 4 1/2 1880 400 7200
B Type N 1 1 6 1410 602 7200
C Type O (Lime Putty) 1 2 9 936 2100 7200
D Type O (Fine Sand) 1 2 9 936 806 7200
E Type O (Coarse Sand) 1 2 9 936 806 7200
F Type O 1 2 9 936 806 7200
G Type K 1 3 12 705 900 7200
H Hydraulic Lime Mortar 0 1 3 0 2640 7200
3.5.2 Mortars for Pier Construction
The mortars used for the construction of test piers were proportioned as were the
mortars for laboratory characterization, except that larger batches were prepared.
For each pier construction, 45.4 kg of sand were used instead of 7.2 kg of sand.
30


These batches were mixed in a large floor mixer. The mortar was re-tempered with
water and mixed with a trowel as required to maintain a workable consistency. Each
mortar batch was used within 2.5 hours of mixing.
31


4 Construction of Test Specimens
4.1 Construction of Piers for Hardness Testing
Photographs of the materials used and the piers under construction can be found in
Appendix C. The piers were constructed in an indoor laboratory on V* in. steel plates
elevated from ground level on bricks. As each course of bricks was added, the
construction was checked for horizontal and vertical level. Six bricks were used for
the perimeter of each course, with no bricks in the center to simulate a cavity wall or
one and a half bricks in the center to simulate a solid wall. Solid piers were built
with full bed, head, and collar joints.
All piers were constructed with Vi in. bed joints. After the piers were constructed
and the mortar partially hardened the joints on two faces of the pier were tooled
concave, while the other two faces were left struck flush with the surface of the
bricks.
With the exception of the hydraulic lime pier, each pier was built with 14 courses of
bricks. The pier built with hydraulic lime contained 12 courses. In some cases, piers
were built with two types of mortar. The piers were allowed to cure in laboratory
air.
Table 4.1 lists each pier with its respective identification number, type of collar joint,
brick type, and mortar type(s).
32


Table 4.1. Pier specifications.
Pier Number Type of Collar Joint Brick Type Dimensions (inches) Number of Courses Mortar Type(s)
1 Cavity Modem Yellow 12 x 20 x 40 14 Type S, play sand
2 Full Reclaimed 13.5x22.5x42 14 Type O, play sand
3 Full Reclaimed 13.5x22.5x42 14 Type O, coarse sand* Type O, fine sand
4 Full Reclaimed 13x22x40 14 Type N, play sand* Type K, play sand
5 Cavity Modem Yellow 12x20x39 14 Type S, play sand
6 Full Reclaimed 13 x 22.5 42.5 14 Type 0, coarse sand
7 Cavity Reclaimed 13x22x44 14 Type 0, Lime Putty and Play Sand
8 Full Reclaimed 13x22x38.5 12 Hydraulic lime, Play sand
*Where two mortar types are given, the bottom 7 courses use the first mortar listed,
and the top 7 courses use the second mortar listed.
4.2 Preparation of Mortar Cubes for Compression
Testing
The mortar cubes were formed as described in ASTM C 109, Standard Test Method
for Compressive Strength of Hydraulic Cement Mortars (using 2-in. or [50 mm]
Cube Specimens). Five sets of three cubes were made from each mortar batch. The
2-in. cubes were formed in brass molds. The mortar was placed in the forms in two
lifts, each tamped in the pattern described in ASTM C 109. After excess mortar was
scraped from the molds with a steel spatula, the forms were covered with a damp
cloth and allowed to cure in the mold for 24 hours. The cubes were then removed
from the molds and stored in a curing chamber until testing. The curing chamber
33


was heated such that the temperature did not fall below 60 Fahrenheit. The relative
humidity in the chamber was monitored and remained between 60 percent and 85
percent. This curing method deviates from the procedure described in ASTM C 109,
which specifies that the mortar cubes are to be cured at 100 percent relative humidity
in a moist room or cabinet. The purpose of this deviation from ASTM C 109 was to
allow more carbon dioxide to reach the specimen and thus encourage the lime in the
mortar to carbonate.
4.3 Preparation of Specimens for Water Vapor
Transmission Testing
Integrated Conservation Resources, Inc. (ICR) of New York, New York provided a
procedure for preparation of water vapor transmission samples. The mortar for these
samples was taken from the batches made for the plastic and hardened mortar tests.
For each mortar type, a set of 3 samples was made. These samples were cured and
then shipped to ICR for water vapor transmission testing according to ASTM E 96,
Standard Test Methods for Water Vapor Transmission of Materials.
34


5 Test Procedures
5.1 Sieve Analysis of Sands
The particle size gradation of the three sand types was determined using ASTM C
136, Standard Test Method for Sieve Analysis of Fine and Coarse Aggregates. Sand
was sifted through a series of progressively smaller sieves and the weight of sand
retained on each sieve was measured. The results, found in section 6, are expressed
as the percent of the total sand weight that passed through each sieve.
5.2 Tests for Mortar Plastic Properties
The following tests were conducted to characterize mortar plastic properties:
Mortar flow, per ASTM C 230, Specifications for Flow Table for Use in Tests of
Hydraulic Cement.
Modified Vicat cone penetration, per ASTM C 780, Test Method for
Preconstruction and Construction Evaluation of Mortars for Plain and
Reinforced Unit Masonry, Annex Al, Consistency by Cone Penetration Method.
Water retention, per ASTM C 110, Test Methods for Physical Testing of
Quicklime, Hydrated Lime, and Limestone.
Air content using the density method, per ASTM C 270.
Water/binder ratio, calculated from the mix proportions used to obtain the
appropriate flow.
Three repetitions of each of these tests were performed on each of the 8 mortar types.
Photographs of these tests in progress can be found in Appendix C.
5.2.1 Mortar Flow
The mortar flow was measured per ASTM C 230. This test was run immediately
after the mortar was mixed. A brass mold and flow table are used for this test.
Mortar is placed into the mold and tamped in two lifts. Excess mortar is then
scraped from the top of the mold. The mold is then swiftly removed, leaving a
35


slumping mortar pile on the table. The table is then dropped from a height of lA in.
25 times by means of a motorized camshaft, causing the mortar to spread into a
pancake shape. Four diameters of this mortar pancake are then measured with
calipers along the lines scribed in the table top.
The sum of four measurements with the calipers is reported as the mortar flow. The
result equals the percent increase in the diameter of the mortar. Samples were mixed
to provide a flow of 110 +/- 5, as required by ASTM C 270 for lab tests. Mixes with
a flow greater than 115 were discarded.
5.2.2 Modified Vicat Cone Penetration
The modified Vicat cone penetration depth was measured per ASTM C 780. This
test was run immediately after the flow test. A Vicat apparatus is used in this test.
First, the dry, empty, brass cup is weighed. Next, the cup is filled with mortar in
three lifts, each lift being tamped 20 times with a 14 in. flat metal spatula. The cup is
then tapped at 5 locations equally spaced around the side of the brass cup with a 5/8
in. diameter, 6 in. long maple dowel. Excess mortar is then cut off to a plane surface
level with the top of the cup with a spatula. The cup and mortar is then weighed.
The tip of the cone is then aligned with the rim of the brass cup, and the cup placed
such that the cone is centered over the mortar surface. The scale is adjusted to read
zero in this condition. The cone is released by a swift turn of its set screw and the
millimeters of penetration are read from the scale.
5.2.3 Water Retention
The water retention of the mortar was measured per ASTM C 110. This test was run
immediately after the modified Vicat cone penetration test. A perforated dish and a
vacuum apparatus are used in this test. Figure 5.1 shows this assembly.
36


# C 110

Ffftr Paper ^_ ,5ito /S6mm-A
f50mn7.otam.-r ^ |
Funnel-'
Three-way
Sfap-cocp-
/ /tier FTcrj/ti
//////d::
IhUjJAz6,:
8JS\SjlsjsY^SyS'SJS\S-75\i.75\ mMmtft
754 fa /56mm.-
\\ * 140 mm. arnm
1 /* 7.7 fo 1.9mm.
rr-Zmm. \| (
f - 3 v..?
7a aspirafor
Mercury
Figure 5.1. Apparatus Assembly for the Water Retention Test (figure taken from
ASTMC110).
A wetted filter is placed in the perforated dish. The dish is filled with mortar and
tamped 15 times. Excess mortar is then scraped from the dish leaving a level
surface. A vacuum of 50.8 mm is then applied to the perforated side of the dish for
60 seconds. After shoveling the mortar about in the dish for 15 seconds, the flow
test as described in Section 5.1.1 is performed using the mortar in the dish. The
water retention is calculated as the flow after suction divided by the flow prior to
suction.
37


5.2.4 Air Content
The air content of the mortar was calculated using the equation for air content in
Section 5.5 of ASTM C 270. The weight of 400 ml of mortar was determined by
subtracting the weight of the 400 ml brass cup from the weight of the brass cup when
filled with tamped mortar. These weights were determined during the Vicat cone
penetration test as described in section 5.2.2.
5.2.5 Water/Binder Ratio
The ratio of water to cementitious binder was calculated from the mix proportions
that were used to obtain the appropriate flow.
5.3 Mortar Hardened Properties
The following tests were conducted to characterize the hardened mortar properties:
Compressive strength, per ASTM C 109, Test Method for Compressive
Strength of Hydraulic Cement Mortars (Using 2-in. or 50-mm Cube
Specimens).
Water vapor transmission, per ASTM E96, Standard Test Methods for Water
Vapor Transmission of Materials.
5.3.1 Mortar Compressive Strength
Mortar cube compressive strength was measured as described in ASTM C 109.
Three 2-in. cubes, formed and cured as described in Section 4.2, were tested for
compressive strength at approximately 7, 14, 28, 60, and 90 days in a hydraulic load
machine. The specimens were loaded to failure and the maximum load and mode of
failure were recorded. The compressive strength was calculated as:
38


Where: fc = compressive strength
P = total maximum load
A = area of loaded surface
The average three tests are reported as the compressive strength of the mortar.
5.3.2 Water Vapor Transmission
The water vapor transmission was measured as described in ASTM E96, Standard
Test Methods for Water Vapor Transmission of Materials. The water method, rather
than the desiccant method, was used. The water vapor transmission is defined as: the
steady water vapor flow in unit time through unit area of a body, normal to specific
parallel surfaces, under specific conditions of temperature and humidity at each
surface. For the water method, the sample is sealed to the top of a dish containing
distilled water. The dish and sample are weighed and then placed in a chamber that
is maintained at a constant relative humidity and temperature. The relative humidity
is maintained at a value of less than 50% with variation of no more than 2%. The
temperature in the chamber is maintained at a value greater than 70 Fahrenheit with
variation of no more than 1 Fahrenheit. Water vapor will pass through the sample
to the chamber. The dish and sample assembly is weighed regularly to record the
amount of water passing through the sample. The water vapor transmission is
calculated using the formula:
WVT
W_
tA
Where: WVT
W
t
A
rate of Water Vapor Transmission
weight of water loss
time between weighings
test area (cup mouth area)
5.4 Rebound Number
An experimental procedure was developed to gather pendulum hammer rebound data
from the laboratory piers. Preliminary data was collected from existing masonry
39


specimens and analyzed. The results of these analyses were used to design the
experimental procedure.
5.4.1 Preliminary Investigation
Existing masonry test walls at the University of Colorado at Boulder Structural
Engineering Laboratory and at the Atkinson-Noland & Associates Laboratory were
used as preliminary test specimens. Grid patterns were established on the faces of
the test walls and pendulum hammer rebound data was taken from 50 or more
locations on each wall. A series of successive rebounds were recorded at each
location without moving the hammer. Analysis of the preliminary data revealed that
at all individual locations the rebound number increased significantly with
successive impacts for the first 3 to 5 impacts, and then settled down to a narrow
range. Figure 5.2 is a plot of a typical series of successive rebound readings from
One location.
Rebound Readings from Succesive Impacts at One
Location
1__________________________________________________________________________________I
Figure 5.2. Rebound readings for a series of successive impacts taken from a test
wall at the University of Colorado at Boulder Structural Engineering Laboratory.
Several different methods were used to interpret the rebound number. These
different methods were analyzed statistically to determine which methods provided
useful results. For example, one method was to take the average of the last 5 of 10
40


impacts of a series as the rebound measurement from that location. Another method
was to take the first rebound number as the rebound measurement from that location.
The methods were analyzed by calculating the variation of the measurements within
the data set and plotting the frequency distribution of the data set. Figure 5.3 shows
the results for a data set with the rebound number interpreted as the average of the
last 5 of a series of 10 impacts. Figure 5.4 shows the results for a data set with the
rebound number taken as the first impact of the series.
Distribution of Rebound Data From CU Laboratory Wall.
Standard Deviation = 5.98, Average = 52.3, Coefficient of
Variation =11.0%, Total # Readings = 53
0.25
Figure 5.3. Results of a statistical analysis of preliminary rebound number data
(rebound number as the average of the last 5 of a series of 10 impacts).
41


Distribution of Rebound Data From OJI aboratorv Wall. Standard
Deviation=4.25, Average=914, Coefficient of Variation =4.6%, Total #
Readings=53
02
S 0.15
(U
If 0.1
u*
b
0.05
0
OOOOOOOOOCnOONOnOO
Rebound Nurrber Average of the highest five readings of each series
Figure 5.4. Results of a statistical analysis of preliminary rebound number data
(rebound number as the first impact of a series).
The coefficient of variation for the first case (using the average of the last 5 of a
series of 10 impacts) is 4.6 percent, while the coefficient of variation for the second
case (using the first impact) is 11.0 percent. This reduction of variation of the data
by more than half is motivation for using a series of impacts to measure the hardness
at one location.
5.4.2 Experimental Data Collection Procedure
The analysis of the preliminary data sets was used to develop a general data
collection procedure for the laboratory study. A numbered grid was drawn on each
pier. Sixty-six grid points were established for both the tooled mortar joints and the
struck mortar joints of each pier.
At each point, the pendulum hammer was positioned to strike the mortar joint and a
reading was recorded from the scale with the impact head resting on the mortar joint
On a perfectly vertical surface, the scale reading is zero in this position. On non-
level surfaces, the scale reading is a positive value or a value below zero. These
negative values were estimated, as the scale does not include negative values. This
42


reading was recorded and later subtracted from the rebound result to correct for non-
level test surfaces.
The piers were tested at approximately 7,14,28, 60, and 90 days of age. Each of
these tests consisted of measurements at 30 or more of the numbered locations. At
each location, a series of 10 or more successive rebounds were recorded. The last 5
of 10 rebounds from each location were averaged to provide one rebound value for
each test location. Then, the 30 or more rebound values were averaged to give one
rebound value for the mortar type, joint type (struck or tooled), and mortar age.
43


6 Results and Discussion
6.1 Sieve Analysis of Sands
Table 6.1 shows the tabular results of the sieve analysis of the three sands used in
this study for construction of test piers. The sieve number gives the number of sieve
wires per inch, the diameter is the size of the largest particle that will pass through
the sieve, and the percent passing is the percentage of the original weight of sand that
passed though a sieve. The gradation curves of these results are plotted below in
Figure 6.1.
Table 6.1. Sieve Analysis Results.
Sieve Number (openings per inch) Diameter (mm) Percent Passing
Fine Sand Play Sand Coarse Sand
4 4.75 100 100 83.1
10 2 99.9 100 72.6
16 1.18 98.5 90.0 61.9
30 0.6 67.0 52.2 42.7
50 0.3 20.7 15.8 15.6
100 0.15 4.72 0.940 4.59
200 0.075 0.161 0.104 0.286
44


Figure 6.1. Sand gradation curves.
6.2 Mortar Plastic Properties
Table 6.2 shows the results of the mortar plastic tests. The results presented are the
arithmetic means of 3 or 4 test results. The percentages in parentheses below the
numerical values are the coefficients of variation of the test results. The coefficient
of variation is defined as the standard deviation of the data divided by the mean.
This value gives an indication of the variability of the results. Low values indicate
low variation.
45


Table 6.2. Mortar Plastic Properties.
Mortar Type Flow Water/Binder Ratio Cone Penetration (mm) Air Content (%) Water Retention (%)
Type S 109 0.68 45.3 7.09 90.1
(Play Sand) (1.6%) (5.6%) (8.7%) (8.0%)
Type N 108 0.84 50.3 5.32 90.5
(Play Sand) (4.6%) (17%) (13%) (0.19%)
Type 0 110 Not 60.7 2.94 93.0
(Lime Putty) (3.8%) Measured (9.4%) (20%) (2.2%)
Type O 110 1.10 55.8 7.80 91.3
(Fine Sand) (2.3%) (4.7%) (8.1%) (4.9%)
Type 0 113 1.02 59.7 5.58 87.6
(Coarse Sand) (2.9%) (9.2%) (3.8%) (3.7%)
Type O 114 0.978 50.7 5.57 89.2
(Play Sand) (1.5%) (4.1%) (4.5%) (4.5%)
TypeK 107 1.12 53.5 6.16 93.7
(Play Sand) (2.6%) (7.2%) (5.0%) (0.9%)
Hydraulic Lime Mortar (Play Sand) 106 (2.7%) 1.28 36.3 (7.9%) = 0 49.9 (2.8%)
Mortar with high lime content has excellent plastic properties, that is, high water
retention and moderate air content. The hydraulic lime has extremely low water
retention, as was observed at the time of mixing. Mortar mixed with this lime
resembled wet sand rather than mortar. This type of mortar may be suitable for
pointing work, but it is not suitable for construction of new masonry. The lime putty
mortar had excellent water retention and was the most workable of all of the mortar
formulations. The low coefficients of variation of this data provide evidence that
these results are accurate.
6.3 Mortar Hardened Properties
6.3.1 Mortar Compressive Strength
The compressive strength of the eight mortar types over time is plotted in Figure 6.2.
Each test result is the average of the measured compressive strength of 3 specimens.
Table 6.3 shows the 90-day and later strengths of the eight mortars.
46


Compressive Strength Over Time of Eight Mortar Formulations
-Type S
-B Type N
-6 Type O (coarse sand)
Type O (play sand)
-ytType O (fine sand)
_ Type O (lime putty)
H TypeK
Hydraulic Lime
Days
Figure 6.2. Compressive strength of mortars over time.
Table 6.3. 90 day and later compressive strength of mortars.
Mortar Type 90 Day and Later Compressive Strength (psi)
Type S 3610
Type N 1660
Type 0 (Lime Putty) 914
Type 0 (Fine Sand) 546
Type 0 (Coarse Sand) 884
Type 0 753
TypeK 354
Hydraulic Lime Mortar 66
47


6.3.2 Water Vapor Transmission
The results of the water vapor transmission tests for the selected mortars are shown
in Table 6.4.
Table 6.4. Rate of water vapor transmission of selected mortars.
Mortar Type Rate of Water Vapor Transmission in [grams/hour]/square meter
Type N (Play Sand) 14.75
Type 0 (Fine Sand) 18.21
Type 0 (Coarse Sand) 19.54
Type 0 (Play Sand) 18.47
Type K (Play Sand) 18.26
Note that these results are for samples of Va inch thickness. These results are not
necessarily indicative of the rate of water vapor transmission through different
mortar thicknesses.
High water vapor transmission rate (WVTR) values indicate high relative
permeability of the material and conversely, low relative WVTR values indicate
lower permeability. Notice that the WVTR values for the 3 type O mortars follow a
trend. The fine sand is the least permeable, followed by the play sand with a higher
WVTR, and then the coarse sand with the highest relative WVTR of the three.
Integrated Conservation Resources Inc. (ICR) compared these results to data in their
company database. Using sands similar to the fine, play, and coarse sands used in
this study, ICR has noted an inverse trend in type O mortars. ICR data shows that
fine sand is the most permeable and the coarse sand is the least permeable. Figure
6.3 shows these opposing trends graphically.
48


Comparison of Water Vapor Transmission Rates for
Different Sand Types
Experimental Samples j
g ICR Data
Figure 6.3. Comparison of water vapor transmission rates of type O mortars with
different sand types.
These contrary results may be due to differences in sample preparation techniques.
The WVTR results for the type N samples show less permeability than type N
mortars in the ICR database. ICR reports that typical results for type N mortars are
18.1 to 21.43 [grams/hour]/square meter, while the type N mortar tested for this
study transmitted only 14.75 [grams/hour]/square meter. This indicates that the type
N mortar is a particularly dense, or non-permeable, type N mortar. The WVTR
results for the type K samples fall within the range of typical values for type K
mortar in the ICR database.
6.4 Rebound Hardness of Laboratory Specimens
Rebound hardness testing was conducted over time with the methods described in
Section 5.4.2. The data was organized in a database and analyzed.
49


6.4.1 Determination of the Number of Tests
Necessary to Characterize Rebound Hardness
Once a large database of rebound data was established, statistical methods were used
to determine the number of measurements necessary to characterize the mortar in an
area of masonry. The methods described in ASTM E 122, Standard Practice for
Choice of Sample Size to Estimate a Measure of Quality for a Lot or a Process, were
used to make this determination.
The following equation from ASTM E 122 was used:
n
' 3 x COV \2
l e
Where: COV = the coefficient of variation of the measurements,
e = percent acceptable error.
3 = a factor corresponding to a low probability (3 in 1000) that
the difference between the sample estimate and the result of
measuring all the units in the lot or process is greater than e. The
choice of the value 3 provides practical certainty that e will not be
exceeded.
n = number of samples required.
The coefficient of variation for use in this equation is that variation that would result
from measurements of all of the members of a lot or all of the measurements of a
process. For the case of masonry mortar rebound data, this conception of the
coefficient of variation is abstract, as it is not possible to measure all of the locations
in an area of masonry. The coefficient of variation was therefore estimated from the
variation calculated from the database of rebound data.
Two sample sizes (representing the number of measurements necessary to
characterize a mortar) were calculated. One sample size was calculated for the
interpretation of the rebound number as the first impact from a location. For this
interpretation, the coefficient of variation calculated from the database was 18.6
percent for 8 groups of 30 or more measurements. It was estimated that, if all of the
locations were measured (instead of 30), the coefficient of variation would be 17
percent. Another sample size was calculated for the interpretation of the rebound
number as the average of the last 5 impacts of a series of 10. For this interpretation,
50


the coefficient of variation calculated from the database was 11.0 percent for 8
groups of 30 or more measurements. It was estimated that, if all of the locations
were measured, the coefficient of variation would be 10 percent.
The number of samples required was then calculated for a range of acceptable errors
for each interpretation of the rebound number. The results of these calculations are
presented in Figures 6.4 and 6.5.
Number of Samples Required as a function of percent
acceptable error. CO.V. = 10% when the average of the last
5 of a series of 10 impacts from each location is used.
e, percent acceptable error
Figure 6.4. Number of samples required when the average of the last 5 of a series of
10 impacts from each location is used.
51


Number of Samples Required as a function of percent
acceptable error. C.O.V. = 17% when only one impact from
each location is used.
e, percent acceptable error
Figure 6.5. Number of samples required when only one rebound is measured from
each location.
Choosing an acceptable error (e) of 10 percent, the number of measurements
required using only one rebound is twenty-six. The number of measurements
required when the last 5 of a series of 10 impacts are used is nine.
6.4.2 Calibration of the Rebound Hammer
A large block of hardened steel was used as a reference material for calibration.
Several series of 10 or more rebounds for the block were used for each calibration
check. The hammer was calibrated soon after it was first received from the
purveyor, and then periodically checked for the duration of the study.
At the midpoint of the study, after more than 5000 rebounds, the arresting brake of
the hammer failed. The arresting brake slides with the pendulum mass along a track
in the frame of the device, and after impact, it arrests the rise of the pendulum at its
apex. The brake wears with use until it no longer has sufficient length to stop the
pendulum from descending after the rebound apex. When thus worn, the brake must
be replaced.
52


After the brake was replaced, the calibration was checked against the steel block, and
it was determined that the rebound from the block was 14 percent less after the repair
than when the hammer was new. The hammer was sent to the calibration lab of the
purveyor for repair and re-certification. However, the rebound reading from the
calibration block remained unchanged after this procedure.
The hammer rebound value from the calibration block was lower than the original
value for the remainder of the study. Efforts to fix the problem by cleaning,
abrading, and polishing the brake track were ineffective. The values from the
calibration block eventually leveled out at 20 percent less than the original value.
6.4.3 Rebound Hardness Development over Time
The rebound number results for selected mortar types over time are shown in Figure
6.6. This data was taken early in the study during a time period when the pendulum
hammer was new and calibration problems had not yet manifested. Later data is
unreliable for comparison over time. These results are of data taken from tooled
joints, using the procedure described in Section 5.4.2.
Figure 6.6. Rebound number versus time for selected mortars.
In all cases, the hardness of the mortar increased as it cured, as would be expected.
However, the increase in hardness does not mirror the increase in compressive
53


strength. For example, the compressive strength of the type N mortar increased by
67 percent from 7 to 28 days of age, while hardness of the type N mortar increased
by 47 percent from 4 to 28 days. This mortar exhibited larger increases in strength
than in hardness during its initial cure. However, the weaker type O and K mortars
demonstrate an opposing trend. These mortars exhibit larger increases in hardness
than in compressive strength during the initial cine. For example, the type O mortar
with fine sand increased in strength by 37 percent from 7 to 28 days, while its
hardness increased by 82 percent. Table 6.5 presents the relative increases in
compressive strength and in rebound hardness of mortars from the first week to the
fourth week of curing.
Table 6.5. Development of compressive strength and rebound hardness of mortars
from the first week to the fourth week of curing.
Mortar Type Percent Increase in Compressive Strength Percent Increase in Rebound Hardness
Type S (Play Sand) 38% 11%
Type N (Play Sand) 67% 47%
Type 0 (Fine Sand) 72% 67%
Type 0 (Coarse Sand) 18% 80%
Type 0 (Play Sand) 37% 82%
Type K (Play Sand) 68% 71%
This trend provides evidence that the rebound number is more sensitive to strength
differences in the lower range of mortar compressive strength (type O and weaker)
and less sensitive to strength differences in the higher range of mortar compressive
strength (type N and stronger).
6.4.4 Hardness of Mortars after 90 Days of Curing
Rebound hardness results for 4 mortar types at an age greater than 90 days and the
compressive strength values of these mortars are presented in Table 6.6. Each data
point is the average of hardness readings from at least 30 points, the reading at each
point taken as the average of the last 5 of a series of 10 impacts. All four of these
mortars contain the same Type 1 cement, Type S lime, and play sand. All data was
taken from tooled mortar joints. The piers containing the type N, O and K mortars
were built with reclaimed Boulder bricks and the pier cavity was filled as
described in section 4.1. The pier containing the type S mortar was built using
modem extruded bricks and the pier cavity was empty.
54


Table 6.6. Rebound hardness of laboratory piers after more than 90 days of curing.
All mortars contain the same cement, lime, and sand.
Mortar Type Hardness of Tooled Mortar Joints (C.O.V) Compressive Strength (psi)
Type S, play sand 81.3 (8.5%) 3610
Type N, play sand 79.9 (5.6%) 1660
Type 0, play sand 70.5 (7.3%) 546
Type K, play sand 47.3 (23%) 354
Rebound hardness results for all 8 mortar types at an age greater than 90 days and
the compressive strength values of these mortars are presented in Table 6.7. Each
data point is the average of hardness readings from 9 points, the reading at each point
taken as the average of the last 5 of a series of 10 impacts. These results are from
data taken after calibration problems with the hammer were noted. For this reason,
these results are not comparable to results reported earlier in this report. However, it
is reasonable to presume, based on subsequent comparative data, that the hammer
calibration remained constant while this data set was collected.
55


Table 6.7. Rebound hardness of tooled and struck joints of all mortar types.
Mortar Type Hardness of tooled mortar joints (C.O.V) Hardness of struck mortar joints (C.O.V.) Compressive Strength (psi)
Type S, play sand1,2 74.1 (8.2%) 64(11%) 3610
Type N, play sand 80.1 (3.6%) 71 (6.8%) 1660
Type O, coarse sand 62.5 (6.7 %) 54.7 (8.2%) 914
Type 0, play sand 76.7 (3.2%) 67.5 (13%) 546
Type 0, fine sand 58.9 (8.5%) 59.6 (5.8%) 884
Type 0, lime putty and play sand1 46.4 (7.9%) 43.8 (13%) 753
Type K, play sand 33.8 (19%) 38.3 (17%) 354
Hydraulic lime, play sand 60.5 (5.6%) 36.8 (26%) 66
'pier constructed as a cavity wall.
2pier constructed with modem, cored bricks
6.4.5 Rebound Hardness versus Compressive
Strength
Figure 6.7 is a plot of the rebound hardness results for the mortars containing the
type S lime, cement, and play sand plotted versus the compressive strength of these
mortars. This data is in tabular form in Table 6.6.
56


Rebound Hardness vs. Compressive Strength: All
mortars contain the same sand, lime and cement. Point
(0,0) is artificial.
Figure 6.7. Rebound hardness versus compressive strength for mortars made with
the same sand, lime and cement.
Notice that there is little difference in the rebound hardness of the type N and type S
mortars although the compressive strength of the type S mortar is more than double
that of the type N mortar. Conversely, there is a significant difference in hardness
between the type K and type O mortars with a similar difference in compressive
strength. Again, the rebound number appears to be more sensitive to changes in
compressive strength in the lower range of mortar compressive strength (type O and
weaker) and less sensitive to strength differences in the higher range of mortar
compressive strength (type N and stronger). However, this observation should be
tempered by the fact that the data for the type S mortar was taken from a pier with an
empty cavity, while the data for the other 3 mortars were taken from piers with full
cavities. The empty cavity could cause the type S rebound hardness to be less than if
the cavity were full.
57


6.4.6 Rebound Hardness versus Joint Tooling
Figure 6.8 is a plot of the rebound hardness results for all of the mortar types used.
This data is in tabular form in Table 6.7.
Figure 6.8. Rebound hardness versus compressive strength for all mortars.
Overall the tooled joints were harder than the struck joints. This result is sensible
given that the tooling process compacts the mortar and thus the joint presents a
denser surface for testing. The type K mortar and the type O mortar with fine sand
were exceptions to this pattern. These exceptions are likely due to differences in the
workmanship of these piers.
This data set was taken late in the study, 6 months after the data presented in Table
6.6 and Figure 6.7. Notice that the results of this later data do not show the same
relationship as between rebound hardness and compressive strength as the earlier,
90-day data. Specifically, the mortars with a higher lime content now show more
hardness relative the mortars with higher cement content. This trend is likely due to
the carbonation of the lime, which is a slower process than the hydration of cement.
With the additional curing time, the lime continued to carbonate while the cement
hydration was essentially complete.
The hydraulic lime mortar, which contains no Portland cement, has very little
compressive strength yet displays rebound hardness comparable to a type O mortar.
58


This phenomenon is noted only in the data from the tooled joints of this pier. This
observation provides evidence that joint tooling has a more significant effect on
rebound hardness than does compressive strength.
Finally, the data from the struck joints shows less variation than the data from the
tooled joints. The average coefficient of variation of data from the tooled joints is
7.8 percent, while the average coefficient of variation of data from the struck joints is
13 percent. This result is sensible given that the tooled joints were carefully tooled
while the struck joints were simply struck flush the trowel. The act of striking the
joints flush leaves a much less consistent joint surface than the tooling procedure.
6.4.7 Rebound Hardness versus Sand Gradation
The rebound hardness results of the type O mortars containing coarse sand, play sand
and fine sand did not show any regular trend that can be attributed to the different
sand gradations.
59


7 Field Tests
Following laboratory development of the test method, a series of field tests were
conducted to determine if the method was viable for in-place evaluation of mortars.
Historic and modem structures were tested. During these tests, attempts were made
to recognize differences in rebound hardness related to different conditions within
one structure, and to catalog the range of mortar hardness results to be expected in
the field.
7.1 Basilica of the Assumption
The Basilica of the Assumption, located in Baltimore, Maryland, was designed by
the prominent architect Benjamin H. Latrobe. Construction of the Roman Catholic
cathedral was undertaken in 1805 and completed in 1837,17 years after the
architects death.
The undercroft, or basement, of the Basilica is a system of masonry vaults. The
masonry is red brick with lime mortar. The lime mortar in this undercroft has
carbonated undisturbed by the elements since its construction.
A series of pendulum hammer tests were conducted on an area in the undercroft.
Figure 7.1 shows the testing in progress.
60


Figure 7.1. Pendulum hammer testing in the undercroft of the Basilica of the
Assumption.
The mortar had developed enough strength to gather data using the method of
repeated impacts. The results showed a rebound hardness of 45, with an error of
13%.
This mortar is the hardest of all the lime mortars measured during this study. This
exceptional hardness is likely due to the length of time that the mortar has
carbonated and the quality of the materials and craftsmanship.
7.2 Saint Alphonsus Church
Saint Alphonsus Church, located in Baltimore, Maryland, was designed by the
architect Robert Cary Long, Jr. Construction was completed in 1845. The exterior
of the church is red brick, and has been pointed with a Portland cement and lime
mortar. Some of the original lime mortar is present in the basement storage rooms of
the church.
Hardness tests were attempted on both of the mortar types. The original lime mortar
was not strong enough to withstand testing. The method of repeated impacts was
61


used on the exterior pointing mortar. This mortar gave a rebound hardness of 54, but
with a 14% coefficient of variation. Also, on several of the impact sites, the mortar
caved into voids behind the pointing mortar. The high coefficient of variation of the
data and the voids behind the pointing mortar indicate that the pointing work was of
low quality.
7.3 Mahan Hall
Mahan Hall was built on the campus of the West Point Military Academy in 1970.
The building is built into rock cliffs above the Hudson River. Figure 7.2 shows the
south elevation of Mahan Hall.
Figure 7.2. South elevation of Mahan Hall, West Point Military Academy.
The exterior mortar joints are cracked and delaminated from the surrounding
stonework. The original construction documents specified that the mortar joints
were to be raked back to 3A inch deep and pointed with a hard, cement-sand mortar.
This pointing mortar was not well compacted into the joints, resulting in a poor bond
to the setting mortar and prevalent voids between the two mortars. The pointing
mortar is debonded and in many cases has fallen out of the joints. Figure 7.3 shows
a joint with both poorly bonded and absent pointing mortar.
62



Portions of the building were repointed with a type N mortar in 1996 to match the
mortar on the interior. This work was done carefully as a test repair. Hardness tests
were conducted on the original bedding mortar, accessed from the interior, and on
the exterior type N pointing mortar. Both mortars were tested using the method for
harder mortar as described in the draft standard test method (Appendix A). The
rebound hardness from measurements of the interior mortar was 50.8 with a
coefficient of variation of 7.2 percent. The rebound hardness from measurements of
the exterior mortar was 65.3 with a coefficient of variation of 19.5 percent.
A petrographic evaluation of the mortars was performed by the Erlin Company of
Latrobe, PA. For the interior mortar, the ratio of cementitious material to sand was
estimated to be 1:3. The mortar was judged to be equivalent to an ASTM C 270 type
N mortar made with masonry cement. The exterior pointing mortar was estimated to
have a ratio of cementitious material to sand of 1:3 to 1:3 Vi. This mortar was
judged to be borderline between an ASTM C 270 type N or S mortar made with
masonry cement, Portland cement, and lime. These results are summarized with the
rebound number results in Table 7.1.
63


Table 7.1. Results of petrographic analysis and rebound hammer testing of mortars
at Mahan Hall.
Rebound Number (C.O.V.) Interior Mortar Exterior Pointing Mortar
51 (7.2%) 65 (19.5%)
Ratio of Cementitious Material to Sand 1:3 1:3 to 1:3 Vi
Estimated Mortar Type TypeN Type N or Type S
The surface of the interior mortar was struck flat with the wall, while the surface of
the exterior mortar was carefully tooled as a test repair. This difference in surface
compaction is likely responsible for the higher magnitude of the rebound number
from the exterior pointing mortar. The high variation, 19.5 percent, of the data from
the exterior is likely due to the test conditions. The exterior of the building is faced
with irregularly shaped granite stones that presented an awkward surface for
pendulum hammer testing.
7.4 Chamberlin Observatory
The Chamberlin Observatory was built in Observatory Park in Denver Colorado in
1890. Figure 7.4 shows the south elevation of the observatory.
64


Figure 7.4. South elevation of the Chamberlin Observatory.
The foundation is stonework with lime mortar. This foundation has suffered damage
from water seeping through the masonry. At lower elevations, the binder appears to
have been leached out of the mortar, leaving a loose, brown sandy material. At
higher elevations in the basement and in the interior telescope foundation, the
original mortar is present in an undamaged state. Several pointing and patching jobs
have been done in this basement with a variety of different mortars. Most of this
work utilized various cement mortars, and one repointing job was done with a lime
mortar containing little or no cement. None of these pointing jobs were
comprehensive, so the mortars of each job are present.
Rebound hardness tests were conducted on the original foundation mortar in its
damaged and undamaged state, on the five pointing mortars, and on original mortar
in the telescope foundation and in an interior brick wall in the cold room of the
observatory.
Two test methods were used. For the softer mortars, the tests were conducted in
accordance with the European standard RILEM MS.D.7: Determination of Pointing
Hardness by Pendulum Hammer, which uses one hit at each test locations. The
harder mortars were tested using multiple impacts at each location. Ten rebound
readings were taken from each test location, and the average of the last five readings

65


was taken as the result from that location. The results of these tests are given in
Table 7.2. Higher rebound numbers indicate a harder, denser material. The rebound
hardness of the original mortar ranges from 10 to 12 for the mortar subjected to
moisture infiltration, to 23 for the undamaged mortar in the telescope foundation.
The rebound hardness of the pointing mortars vary widely, varying from 15 to 79.
Table 7.2. Rebound number test results from the Chamberlin Observatory.
Mortar location and description Method* Rebound Number
Original mortar, interior foundation, deteriorated lower courses on north wall A 10
Original mortar, interior foundation, intact mortar at upper courses on south wall A 12
Original mortar, telescope foundation, east side A 23
Original mortar, interior of coal room brick wall A 20
Pointing mortar (soft, light-grey colored), prevalent on northeast of foundation A 15
Pointing mortar (dark grey), used A 50
extensively, tested at lower courses on east side B 67
Pointing mortar, (light grey, slightly A 35
glossy), used extensively above heavy efflorescence on northeast side B 46
Pointing mortar, (crumbly dark grey), A 18
tested on west side B 25
Pointing mortar, cement patches A 65
prevalent throughout south side B 79
*Method A: Value is the median of initial rebound number at 9 different location
(RILEM MS.D.7).
*Method B: 10 rebound hits recorded, greatest 5 values are averaged; the final
rebound number is the average for either 4 or 5 separate locations.
The rebound hardness results from the Chamberlin Observatory clearly differentiated
between the various repair mortars. Some of the pointing mortars were
inappropriately hard for the conditions, and the rebound results for these mortars
proved to be very high compared to the original, undamaged mortar. Some of the
66


pointing mortars were more carefully matched, and the rebound results from these
mortars were closer to the results from the original, undamaged mortar.
7.5 The Graham Bible House Carriage House
The carriage house of the Graham Bible House is located in City Park in Denver,
Colorado. This turn of the:century construction is a registered landmark (Landmark
No. 221) with the City and County of Denver Landmark Preservation Commission.
Figure 7.5 shows the carriage house from the southwest.
Figure 7.5. Southwest elevation of the Graham Bible House Carriage House.
The carriage house masonry walls are soft molded brick with lime mortar. The
exterior is painted. The masonry is decayed, largely due to damage from lawn
sprinklers. Pendulum hammer hardness tests were conducted on the north and south
walls. The north wall was painted, and the paint was removed from the test
locations. The paint on the south wall had fallen off from the test locations, which
were in the path of the sprinkler water spray.
Both locations were tested using the method for harder mortar as described in the
draft standard (Appendix A). The rebound hardness result from measurements of the
north wall was 32.4 with a coefficient of variation of 27 percent. The rebound
67


hardness result from measurements of the south wall was 35.0 with a coefficient of
variation of 13.2 percent.
It was noted that the mortar on the north wall consisted of a hard, external layer to a
depth of 1/8 inch to 3/16 inch, and much softer mortar beneath that harder shell. It is
likely that the wall had been repointed. On a few occasions the pendulum hammer
impact penetrated the hard exterior after several impacts, and then subsequent
impacts gave lower values. These occurrences help to account for the large
coefficient of variation of the data set taken from this wall.
On the south wall, the extent of deterioration of many regions was such that those
regions could not be tested. The regions that were intact, however, demonstrated
hardness values that can be considered healthy for a lime mortar. Thus these results
may give a false impression of the hardness of the mortar, as the weaker mortar has
been washed away, leaving only the sound material.
7.6 B. Jr.s Auto Parts Store
B. Jr.s Auto Parts store of Casper, Wyoming, is a masonry building constructed of
nominal 8 by 8 by 16 concrete masonry units with a rectangular plan of
approximately 100 by 35. The building sustained fire damage on July 21,1998.
Pendulum hammer rebound testing was conducted on damaged and undamaged
mortar in the building. Table 7.3 shows the results of the pendulum hammer testing.
Table 7.3. Pendulum hammer test results from B. Jr.s Auto Parts Store.
Condition Mean Coefficient of Variation Number of Samples
Fire Exposed 48 6.9% 25
Fire Sheltered 62 5.0% 25
These results show that the fire had reduced the hardness of the mortar. The
decrease in hardness implies a decrease in strength. The rebound hardness method
was used to identify damaged regions requiring repair. This case is an example of
another useful application of pendulum hammer mortar testing. The problem in this
case, the evaluation of comparative deterioration within one structure, is well
addressed with pendulum hammer rebound testing.
68


7.7 The Centennial School Gymnasium Addition
Centennial Elementary School is located in Colorado Springs, Colorado. The
addition of a new gymnasium was completed in the summer of 1998. The results of
compression tests of mortar and prism specimens cast during construction showed
that some of these specimens had less than the allowable strength specified by the
project engineer. Pendulum hammer hardness tests were performed in conjunction
with petrographic analysis to evaluate the mortar.
Three mortar samples were removed from the building for petrographic analysis.
The sampled areas were also tested for hardness with the pendulum hammer.
Further hardness testing was done on other areas in order to determine if the three
sampled areas were representative of the mortar throughout the structure. The
testing was conducted using the method for harder mortar as described in the draft
standard (Appendix A).
Three wall types were present and tested at several locations. The three wall types
tested were a brick veneer, a concrete block wall, and a composite, grouted concrete
block and brick wall that was tested from the brick side. The highest rebound
magnitude was recorded from tests of the composite wall, which is very rigid and
thus unlikely to be moved by the energy of the hammer impact. A sample was
removed from this wall for petrographic analysis. The petrographer estimated that
the original mix proportions of this mortar were 1 part cement, Yz part lime, and 7 Yz
parts sand by volume. These proportions describe an over-sanded type S mortar.
The lowest rebound values were recorded from mortar between ungrouted concrete
masonry units, which present a less rigid surface for impact resistance. The
petrographer estimated that the original mix proportions of a mortar sample from this
wall were 1 part cement, Yz part lime, and 4 Yz parts sand, which describes a type S
mortar. The results from the brick veneer, which is more rigid than the concrete
masonry units but less rigid than the composite, grouted wall, showed rebound
magnitudes in-between the other two cases. The petrographer estimated that the
original mix proportions of a mortar sample taken from this wall were also those of a
type S mortar. Thus the results of hardness testing did not correlate to the
petrographic mortar analysis. It would be reasonable to expect that the over-sanded
mortar would give a lower rebound value, but instead this mortar gave the highest
rebound value. Thus it is likely that in this case the type of wall construction had a
greater influence on the rebound number than did the mortar quality. Table 7.4
summarizes these findings.
69


Table 7.4. Hardness results from the three types of wall construction at the
Centennial School Gymnasium Addition.
Type of wall construction Relative rigidity Average rebound number Estimated mortar mix proportions from petrographic analysis (cement: sand: lime)
Composite brick, block, and grout. Very rigid 79.3 1: Vi-.lVi
Brick veneer with insulated cavity Less rigid 73.2 1: Vi: 4 Vi
Ungrouted concrete masonry unit Least rigid 62.7 1: l/2:4!/2
This case is an example of another useful application of pendulum hammer mortar
testing. The pendulum hammer testing provided evidence that the mortar samples
taken for petrographic analysis were representative of the mortar in the whole wall
rather than just the sampled area. This is important evidence because petrographic
analyses are very expensive and somewhat destructive.
70


8 Conclusions and Recommendations
Mortar hardness testing using the pendulum hammer is a quick, nondestructive
method that is useful for the evaluation of in-place mortars. Mortar hardness
measurements were conducted on laboratory specimens, historic and recent masonry
structures, a fire-damaged structure, and on a newly constructed structure. Both
original mortar and pointing mortar were tested. A test method was developed after
analysis of these measurements. Rebound hammer testing proved useful for many
evaluation applications. Problems encountered during the study are discussed here
and solutions are recommended.
8.1 Factors Affecting Rebound Hardness
The effects of mortar type, compressive .strength, sand type, joint tooling, and curing
time on mortar rebound hardness were investigated. It was found that rebound
hardness was dependent on mortar type. Mortar types with higher cement content,
and thus higher compressive strength, tended to be harder than those containing less
cement. Mortar joints which were tooled tended to be harder than those which were
struck flush with the brick surface. Mortar hardness also increased with additional
curing time. No correlation was noted between sand gradation and rebound
hardness.
It was noticed that rebound hardness was affected by the rigidity of the masonry
construction. Mortar in rigid construction such as composite, grouted brick and
block walls give higher rebound numbers than mortar in less rigid construction such
as facing brick in a cavity wall.
8.2 Development of Test Methods
Two methods of rebound hardness testing were developed. These methods were
designed based on statistical analysis of the database of rebound hardness results
collected from the laboratory specimens. The variability of the data was used to
choose the number of tests required to produce statistically significant results. One
method, designed for hard mortars, is to impact 9 locations in an area of masonry 10
times in succession. The rebound hardness at each of these 9 locations is taken as
the average of the last 5 of the 10 hits, and the rebound hardness for the masonry
71


area is taken as the average of the rebound hardness from the 9 locations. The other
method, designed for soft mortars, is to impact 26 locations in an area of masonry
once at each location. The mortar rebound hardness for the masonry area is taken as
the average of these 26 impacts. A test method has been written and submitted to
ASTM for consideration as a standard test method. This method can be found in
Appendix A of this report.
8.3 Usefulness of Rebound Hardness Results
The method proved useful for comparing deterioration, damage, or quality of mortar
within a structure. The method was used to delineate fire-damaged regions of a
structure. The rebound hardness of the fire-damaged regions was less than the
hardness of undamaged regions. The results were used to specify which sections of
the building required repair. Another application was the investigation of mortar
quality of a recently constructed structure. Samples were removed for petrographic
analysis, and rebound hardness measurements were used to compare the sampled
areas to the rest of the structure. In this way it was possible to generalize the results
of the petrographic analyses of the local samples. In several field trials, poor quality
pointing was identified when the hammer fractured the pointing mortar, revealing
voids behind the pointing mortar. In other applications, it was possible to determine
whether pointed mortar matched the original mortar of a structure. These examples
emphasize the usefulness of comparative mortar hardness results from within one
structure. The use of quantitative results alone is discouraged, except for the
determination of an appropriate range of pointing mortar strength. The results of
rebound hammer testing are precise enough to be used for the type-matching of
repointing mortar to original mortar.
8.4 Problems and Limitations
Problems encountered during rebound hammer testing include calibration difficulties
and the need to frequently repair the hammer, joint depth and thickness limitations,
removal of surface coatings, and the dependency of the results on the mass and
stiffness of the element tested.
During this study, the hammer calibration reference values changed with time. The
hammer gave lower readings from the calibration block with extended use. This was
likely due to roughening of the hammer brake slide. Also, it was necessary to
replace the hammer brake periodically. It is possible that these problems were due to
72


malfunctioning of the particular hammer used, and that other hammers of the same
model do not have this problem.
Rebound hammer testing was not possible on masonry with joint thickness of less
than 3/8 inch, nor on masonry with joint depth of more than 3/8 inch. The depth
limitation was problematic during field testing of masonry with deteriorated mortar.
The effect of mass and stiffness of the masonry element tested inhibits the
comparison of rebound hardness data from different structures. In addition, when
testing a single structure, care must be taken to avoid comparing results from regions
of different stiffness.
8.5 Future Work and Recommendations
A database of rebound hardness results of mortar types from around the country
would expedite the use of quantitative hardness results. Calibration difficulties will
have to be addressed prior to this undertaking. The beginnings of such a database
have been produced during this study.
The quantification of the effects of the stiffness and mass of the specimen tested on
the rebound hardness would be a useful addition to the test method. By measuring
the rebound hardness of masonry specimens of different thickness and measuring the
vibrations of the specimens, guidelines could be created to remove these effects from
the results.
The brake mechanism of the hammer is the cause of the calibration difficulties
encountered in this study. Replacing this mechanism with another may produce
more reliable results. Possible replacements include a ratchet mechanism and a slave
indicator needle.
The limitations of joint thickness and depth could be overcome with the use of
different impact heads. As equipped, the impact head is not replaceable. An
assortment of interchangeable impact heads and pendulum masses could be
developed such that all mortar joints would be accessible.
73


Appendix


Appendix A. Draft Standard Test Method
Standard Test Method for the Determination of the Rebound Hardness of
Masonry Mortar
1. Scope
1.1 This test method covers the determination of the rebound number of hardened
mortars using a pendulum hammer with a steel impact head. The test method
is limited to masonry mortar with joint thickness greater than or equal to 3/8
in. (9.53 mm) that are accessible for testing with the rebound hammer shown
in figure 1.
1.2 Use of this test method may cause 3/8 in. (9.53 mm) diameter circular
indentations in mortar joints up to 5/16 in. (7.94 mm) deep.
1.3 This standard does not purport to address all of the safety problems, if any,
associated with its use. It is the responsibility of the user of this standard to
establish appropriate safety and health practices and determine the
applicability of regulatory limitations prior to use.
2. Referenced Documents
2.1 ASTM Standards:
E 177 Practice for Use of the terms Precision and Bias in ASTM test methods1
E 122 Standard Practice for Choice of Sample Size to Estimate a Measure of
Quality
For a Lot or a Process2
3. Significance and Use
3.1 The purpose of this test method is to determine a hardness property of
masonry mortar in-situ with a simple apparatus. The rebound number has no
units and describes the magnitude of the rebound of a mass from a surface
when released from a certain height. This test is suitable for use as an aid in
the evaluation of masonry and the selection of pointing mortar.
4. Apparatus
4.1 Pendulum Rebound HammerThe pendulum rebound hammer shall be the
Schmidt type PM pendulum hammer manufactured by the Proceq Company
of Switzerland. This hammer is shown in Fig. 1.
1 Annual Book of ASTM Standards, Vol. 14.02
2 Annual Book of ASTM Standards, Vol. 14.02
75


Fig. A. 1 Pendulum Rebound Hammer
4.2 Calibrate the pendulum rebound hammer using a steel anvil of the type
available from the manufacturer or another such mass that gives consistent
rebound numbers. It is recommended that the pendulum hammer be
calibrated after 1000 impacts. The maximum allowable deviation from the
calibrated reference value is +/- 2 units. The rebound values are sensitive to
temperature variations of the calibration mass and the rebound hammer.
5. Test Area
5.1 Selection of Test AreaMasonry shall be fixed within a structure or rigidly
supported. The surface of the mortar joint must be recessed no more than
5/16 in. (7.94 mm) from the surface of the masonry units. The joint thickness
76


must be 3/8 in. (9.53 mm) or more to accommodate the impact head. T ^
masonry surface must be near vertical. Tooled joints yield higher reboi ind
numbers than struck joints.
5.2 Surface PreparationPaint and other sealants must be removed from tjie
mortar surface prior to testing.
6. Procedure for Hard Mortars
6.1 For this procedure, a hard mortar is defined as one that, after being impacted
by 10 successive rebound hammer blows on the same location and in tl e
same direction, will be indented a distance less than that distance whicq will
cause the range of motion of the hammer to be exceeded.
6.2 Position the rebound hammer such that the hammer head will fall in a Vertical
plane and strike a mortar bed joint.
6.3 Prior to impacting the surface, allow the hammer head to rest on the mdrtar
and record the reading from the scale as the initial offset.
6.4 While holding the device firmly in place, impact the mortar surface ten times
and record the rebound numbers.
6.5 Repeat this procedure at 9 or more locations in a region of interest that
appears to have uniform properties3.
7. Procedure for Soft Mortars
7.1 For this procedure, a soft mortar is defined as one that, after being imp* qted
by 10 successive rebound hammer blows on the same location and in th
same direction, will be indented a distance greater than or equal to that
distance which will cause the range of motion of the hammer to be exc< eded.
7.2 Duplicate the procedure for hard mortars but impact the surface once instead
of ten times and at 26 locations instead of 9 locations within the test arc a3
8. Calculation of Results
8.1 Calculation of Average Rebound Number for Hard MortarsFor each
of 10 rebound readings from one location, discard the first 5 of the 10
readings. Subtract the initial offset recorded for that location from eacl
the 5 remaining readings. Calculate the arithmetic mean of all of the
3 The number of impact locations was determined using ASTM 122, Standard Practice for Ch
Sample Size to Estimate a Measure of Quality for a Lot or a Process, using a maximum allow;
sampling error of 10% of the process mean, and a probability factor of 3 to give a 3 in 1000
probability that the sampling error will exceed the maximum allowable sampling error. For h; ri
mortars, a coefficient of variation of 10% was used. For soft mortars, a coefficient of variatioi. }f 17
% was used. These coefficients of variation were determined by calculating the arithmetic me i: i of
the coefficients of variation of 8 groups of more than 30 impacts or series of impacts performe d on
laboratory test specimens.
enes
Df
ce of
ble
77


remaining corrected readings for the 9 or more series of impacts from the test
area.
8.2 Calculation of Average Rebound Number for Soft MortarsSubtract the
initial offset recorded at each location from the rebound reading from that
location. Calculate the arithmetic mean of the 26 or more corrected readings
from the test area.
9. Report
9.1 The report shall include the following:
9.1.1 The name of the person or persons conducting the tests.
9.1.2 The hammer type, serial number, and evidence of calibration.
9.1.3 Identification of the structure tested and date of construction, if
available.
9.1.4 Description of testing conditions (for example, temperature).
9.1.5 Location of each area tested within the structure.
9.1.6 Description of the masonry:
9.1.6.1 Type of masonry unit, including any known material properties
and any known history.
9.1.6.2 Type of mortar, including any known material properties or
constituents, any known history including age and the date of any
pointing work and the type of pointing mortar.
9.1.6.3 Sketch of wall configuration, including joint thicknesses, number
of wythes, cavity or collar joint, and wall thickness.
9.1.6.4 Test procedure used (procedure for hard or soft mortar) and the
average rebound number for each test area.
10. Precision and Bias
10.1 Precision
10.2 The available test data shows the coefficient of variation of this test method
to be as great as 10% and it is recommended that a minimum of three tests
be conducted in the same general area to verify test results.
11. Keywords
in-situ; hammer; hardness; masonry; mortar; nondestructive evaluation; pendulum
78


Appendix B. Referenced ASTM Standards
ASTM C 109, Standard Test Method for Compressive Strength of Hydraulic Cement
Mortars (using 2-in. or [50 mm] Cube Specimens
ASTM C110, Test Methods for Physical Testing of Quicklime, Hydrated Lime, and
Limestone
ASTM C 136, Standard Test Method for Sieve Analysis of Fine and Coarse
Aggregates
ASTM C 150, Standard Specification for Portland Cement
ASTM C 207, Standard Specification for Hydrated Lime for Masonry Purposes
ASTM C 230, Specifications for Flow Table for Use in Tests of Hydraulic Cement
ASTM C 270, Standard Specification for Mortar for Unit Masonry
ASTM C 780, Test Method for Preconstruction and Construction Evaluation of
Mortars for Plain and Reinforced Unit Masonry, Annex Al, Consistency by Cone
Penetration Method
ASTM C 805, Test Method for Rebound Number of Hardened Concrete
ASTM C 1072, Test Method for Measurement of Masonry Flexural Bond Strength
ASTM C 1324, Standard Test Method for Examination and Analysis of Hardened
Masonry Mortar
ASTM E96, Standard Test Methods for Water Vapor Transmission of Materials
ASTM E 122, Standard Practice for Choice of Sample Size to Estimate a Measure of
Quality for a Lot or a Process
79


Appendix C. Ancillary Photographs
80


Figure C.2. From left to right: fine sand, play sand, and coarse sand.
81


Figure C.5 Steel plate with mortar base prepared for pier construction.
82


Figure C.6. Checking for horizontal level during pier construction.
Figure C.7. A pier in construction with a solid collar joint.
83


84


Figure C. 10. Mortar cubes and a brass mold
Figure C. 11. Mortar is tamped in the brass mold on the flow table.
85


Figure C. 12. Excess mortar is scraped from the top of the mold to leave a level
surface.
86


Figure C. 14. The diameter of the mortar pancake is measured with calipers.
87