Clemson University TigerPrints
All Theses Theses
12-2012 Characterization and restoration of historic Rosendale cement mortars for the purpose of restoration Stephanie Hart Clemson University, [email protected]
Follow this and additional works at: https://tigerprints.clemson.edu/all_theses Part of the Materials Science and Engineering Commons
Recommended Citation Hart, Stephanie, "Characterization and restoration of historic Rosendale cement mortars for the purpose of restoration" (2012). All Theses. 1549. https://tigerprints.clemson.edu/all_theses/1549
This Thesis is brought to you for free and open access by the Theses at TigerPrints. It has been accepted for inclusion in All Theses by an authorized administrator of TigerPrints. For more information, please contact [email protected].
CHARACTERIZATION AND RESTORATION OF HISTORIC ROSENDALE CEMENT MORTARS FOR THE PURPOSE OF RESTORATION
A Thesis Presented to the Graduate School of Clemson University
In Partial Fulfillment of the Requirements for the Degree Master of Science Materials Science and Engineering
by Stephanie Anne Hart December 2012
Accepted by: Dr. Denis Brosnan, Committee Chair Dr. John Sanders Dr. Kathleen Richardson
ABSTRACT
Mortar was a very common building material in today’s historic sites. Before Portland
cement was manufactured at a global level, Rosendale cement was commonly used in
these mortars. Over time, these mortars in historic sites have begun to break down and wear away. With Rosendale cement in production again, measures can be taken to restore and repair the historic mortars. However, little testing has been done to establish durability of modern Rosendale cement mortars. This presentation highlights the common mix techniques used at the time, and undergoes experiments to establish general properties and predict future durability. Six different mortar mixes were tested with varying cement content and using various lime additions. Properties observed include compressive strength, absorption, porosity, permeability, and bond strength. Ion chromatography was used on seawater-soaked samples to determine how the Rosendale cement mortar would react with the seawater. Relationships between these properties were also addressed. It was found that cement content played a large role in compressive strength, while lime content had an effect on bond strength. Ion chromatography was used on seawater-soaked samples to determine how the Rosendale cement mortar would react with the seawater. Magnesium sulfates, and chloride were taken up into the mortars, indicating that Rosendale would be venerable to salt attack.
ii DEDICATION
I would like to thank my family, close friends, and my love Graham. Without all of you,
I would have surely given up or gone crazy. Your belief and support in me means
everything.
iii ACKNOWLEDGMENTS
There were many people involved that made this thesis possible. First, I would like to thank my advisor, Dr. Brosnan for his guidance and direction. I would also like to thank
Dr. Sanders, for his help and advice with all my testing and Dr. Richardson for her
participation in my committee from afar.
None of this would have been possible without Clemson University for accepting me into
the program and the National Park Service for the funding of this project.
Finally, I would like to thank everyone at the National Brick Research Center: Dr. John
Sanders, Gary Parker, Michael Mason, Graham Shepherd, Parker Stroble, Mark Young, and Greg Bellotte; for providing the facilities, testing equipment, and their continued help and support throughout my experience.
iv TABLE OF CONTENTS
Page
TITLE PAGE ...... i
ABSTRACT ...... ii
DEDICATION ...... iii
ACKNOWLEDGMENTS ...... iv
LIST OF TABLES ...... vii
LIST OF FIGURES ...... viii
CHAPTER
I. INTRODUCTION AND MOTIVATIONS ...... 1 Introduction ...... 1 Motivations ...... 2
II. LITERATURE REVIEW ...... 4
Rosendale Cement ...... 4 Strength ...... 9 Salt Attack - Introduction...... 11 Sulfate Attack...... 10 Magnesium Effects ...... 15 Effect of Chlorine ...... 18
III. EXPERIMENTAL PROCEDURES ...... 23
Description of Materials and Mixes ...... 23 Fabrication of Mortars ...... 24 Compressive Strength Test ...... 27 Tensile Splitting ...... 29 Bond Wrench ...... 30 Pier Compression ...... 31 Water Vapor Transmission Test ...... 34 X-Ray Diffraction ...... 35
v Table of Contents (Continued)
Page Thermogravimetry and Differential Scanning Calorimetry ...... 35 Mercury Intrusion Porosimetry ...... 36 Initial Rate of Absorption ...... 36 X-Ray Flouresence...... 36 Salt Leaching Experiments ...... 37
IV. RESULTS AND DISCUSSION ...... 38
Rosendale Characterization ...... 38 X-Ray Diffraction Mix Characterization ...... 40 X-Ray Fluorescence Characterization ...... 41 Simultaneous Thermal Analysis ...... 43 Porosity ...... 46 Compressive Strength ...... 48 Water Vapor Transmission ...... 52 Initial Rate of Water Absorption ...... 54 Summary of Mechanical Testing ...... 56 Bond Strength ...... 57 Pier Compression ...... 58 Magnesium Sulfate Attack ...... 58 Chloride Effect ...... 66
V. CONCLUSIONS...... 69
APPENDIX A: RAW DATA ...... 71
REFERENCES ...... 81
vi LIST OF TABLES
Table Page
2.1 Chemical compositions of various natural Roman and American cements ...... 5
2.2 Phase composition of various natural Roman and American cements ...... 6
2.3 Typical composition of seawater ...... 11
3.1 Description of mixes used in experiments in weight percent ...... 24
4.1 Oxidized chemistry of raw materials ...... 39
4.2 Oxidized chemistry of mortar mixes ...... 42
4.3 Loss of ignition for the tested mixes ...... 42
4.4 MIP results for all tested mixes ...... 46
4.5 Summary of results sorted by mix ...... 56
4.6 Results of the bond strength test ...... 57
vii LIST OF FIGURES
Figure Page
2.1 Micrograph of clinkered Portland cement ...... 7
2.2 Graph showing the relationship of porosity with strength ...... 10
2.3 DTG of Portland cement paste hydrated for one week ...... 14
2.4 TG and DTG of a plain Portland cement mortar (a) before and (b) after 200-day exposure to 5% MgS solution ...... 17
2.5 Expansion of Portland cement mortars stored in seawater and groundwater solutions ...... 19
2.6 Strength reduction in Portland cement mortars ...... 20
3.1 Addition of water to dry mortar mix ...... 25
3.2 Placement of the wet mortar mix into the molds ...... 26
3.3 Fresh mortar cubes in the initial humidity chamber ...... 27
3.4 Compressive strength testing of the 2”x2” mortar cubes ...... 29
3.5 A brick couplet in the bond wrench apparatus after failure ...... 31
3.6 A brick pier following the pier compression test ...... 33
3.7 Water vapor transmission testing apparatus ...... 34
4.1 X-ray diffraction pattern of Mix C ...... 41
4.2 Mix C TG, DSC, water, and carbon dioxide analysis after 28 days curing ...... 44
4.3 Thermogravimetry of all tested mixes ...... 45
viii List of Figures (Continued)
Figure Page
4.4 Comparison of cement content with porosity with different mix types ...... 47
4.5 Comparison of different mix types’ cement contents with bulk density ...... 48
4.6 Compressive strength over 28 days of curing ...... 49
4.7 Comparison of cement content and compressive strength ...... 51
4.8 Comparison of porosity and strength in tested mixes ...... 50
4.9 Results of the water vapor transmission test for each mix ...... 53
4.10 Comparison of water vapor transmission and volume of large pores...... 54
4.11 Water absorption for the mortar mixes over a 24-hour period ...... 55
4.12 Sulfate ion concentration in seawater over a 180-day soak period ...... 59
4.13 TG, derivative TG and water analysis of an unsoaked Mix C cube ...... 60
4.14 Mix C TG, DTG, and water analysis after the 180-day seawater soak ...... 61
4.15 Comparison of Mix C unsoaked (green) and soaked (red) TG/DTG and water analysis ...... 62
4.16 Magnesium ion concentration in seawater over a 180-day soak period ...... 63
4.17 Concentration of magnesium ions varying lime putty content...... 64
ix List of Figures (Continued)
Figure Page
4.18 Concentration of calcium ions varying lime putty content...... 66
4.19 Chloride ion concentration in seawater over a 180-day soak period ...... 67
x CHAPTER 1
INTRODUCTION AND MOTIVATIONS
Introduction
Cement has long been an important material in the use of construction. Its first uses can
be dated back over 4000 years ago in Egyptian times [1]. Brick and mortar were also a
key component in building America. Following the Revolutionary War and the War of
1812, natural cement was primarily used in mortars to build buildings and forts. This use
of natural cement continued through the Industrial Revolution. In 1875 the first
American Portland cement was produced. By the turn of the century, Portland cement
dominated production.
During the time of natural cement, a majority of it came from cement rock mined in
Ulster County, New York, near a town called Rosendale. Nearly half of all natural
cement used was Rosendale [1].
Thinking back, the structures built with natural cement are over a century old. Some of
these structures are beginning to degrade, so interest in natural cement has resurfaced in
the interest of restoration. It is usually recommended that a repair mortar is of similar compositions and physical properties [2]. Because Rosendale cement was the most
popular at the time, it has again become available for these restoration projects. Because
of its popularity historically, this project will focus on using modern Rosendale cement
for fabrication of the mortars.
Over a mortar’s lifetime, various factors affect how the mortar will perform. Some of these factors are apparently evident, such as strength and porosity. Others take time to make themselves apparent, such as resistance to salt attack or freeze/thaw durability.
However, these factors could also be estimated using analysis techniques such as X-ray
fluorescence, X-ray diffraction, and permeability.
Understanding salt attack is essential for restorative mortars. One of the main issues
today in historical monuments is salt crystallization [3]. Many structures were built on or
near coastlines. Knowing how the seawater compromised the mortar can lead to a better understanding on how to make mortars to minimize this effect.
The goal of this project was to characterize modern Rosendale mortars and establish a
relationship between composition and durability. Being able to predict how the mortars
will react will lead to a better understanding of how to use Rosendale cement in
restoration for the future.
Motivations
The primary motivation for this project was to determine if modern Rosendale cement is
a reasonable material to use for restoration projects in structures that used Rosendale
2 cement at the time of construction. The goal was to improve the structure while not
causing any harm to the existing components. The modern Rosendale mortar must therefore have a high compressive strength, good durability from water and seawater ions, and not cause any future damage, such as calcium leaching out to damage bricks
[3]. Cement is the most expensive component of a mortar, so retaining good properties while minimizing the cement and cost was also important. This project looked at varying cement contents, plus some cement replacement with other lime-based binders. It was investigated to see how the binders would act and if they would be an option to use in restorations.
3 CHAPTER 2
LITERATURE REVIEW
Rosendale Cement
Natural cement was produced and used in construction from 1818 through 1970. Of this
time, over half of the 35 million tons produced originated from near the town of
Rosendale, NY [4]. Natural cement is made from the natural cement rock that is
excavated. These “cement rocks” are limestone rocks rich in clays [5]. More
specifically, Rosendale is produced from argillaceous sedimentary rocks with high
dolomitic content. The rock is then calcined by burning, and then ground into a powder
to form the cement.
Natural cement differs from lime cement in two noticeable ways. First, it does not slake,
or become liquid, when water is added. Also, it demonstrates hydraulic properties such
as setting under water [6]. Even though it exhibits hydraulic properties, it still differs
from the most popular hydraulic cement today – Portland cement. Portland cement is a
chemically controlled, synthetic mixture that has little variability in chemistry or
properties. Natural cement comes from burnt natural rock, which can show significant
variability between layers and locations. Variation in color and chemistry are common.
Table 2.1 shows chemistries of some Rosendale cements. A physical property that
differs greatly from Portland cement is the time it takes to set, or harden. Natural cement
sets much more quickly than Portland, resulting in a lower ultimate strength [6].
4
Table 2.1: Chemical compositions of various natural Roman and American cements [5]
Chemically, Rosendale cements can be very comparable to some Portland cements.
Table 2.1 shows the variation between numerous natural cements, including two different
Rosendale mixes. Even if two cements have identical chemistries, the difference in cement lies within the firing of the rock. Portland ceement is fired in excess of 1400oC, commonly called “clinkering.” The calcium silicate phase alite (C3S) is formed at around
o 1300 C from belite (C2S) and lime. This presence of alite is what causes the differences in strength and set time. Rosendale cement was produced when tthese temperatures could not be obtained, so no alite forms. Rosendale cement is; however, fired over the calcination temperature of around 800oC, which lends to the cement having hydraulic properties [7]. This is when the belite phase forms, which is the cement phase that hydrates to form the calcium-silicate-hydrate (C-S-H) gel phase, which gives the cement its strength.
5 Table 2.2 shows the phase compositions for the same cements observed in Table 2.1. As expected, the Rosendale mixes did not contain any alite. It is important to note that the
Rosendale mixes contained significant quantities of periclase, or crystalline MgO, compared to other natural cements. This can also be used as an identifying tool to distinguish Rosendale forensically from other natural cements. This is due to the preexisting limestone already present in the Rosendaale cement rocks. This does not cause much of a problem in America; however, since Rosenndale was one of the predominant natural cements used.
Table 2.2: Phase composition of various natural Roman and American cements [5]
Figure 2.1 shows a micrograph of a cement with alite present, indicating that the cement was clinkered and is, therefore; not a natural cement..
6
Figure 2.1: Micrograph of clinkered Portland cement [7]
Both lime paste and powdered lime hydrates were commonly added to Rosendale cement
[4]. Lime was a common additive because it added favorable workability for the mason, demonstrated some self-healing properties, and increased the bond strength between mortar and brick [8]. The joint strength and self-healing comes as a result of the chemical reaction:
Ca(OH)2 + H2O + CO2 CaCO3 + 2 H2O
This calcium carbonate can fill small cracks or voids in the mortar itself, or those within the adjoining brick.
7 Sand was also added as aggregate to complete the mortar mixture. Aggregates are used to give the mortar volume, as cement paste is expensive alone. Based on the application, binder-sand ratios varied from 1:1 to 1:2¾ [6].
With the reintroduction of Rosendale cement in 2004, restoration projects that were once overlooked can now be attainable [4]. Many structures in the United States contain
Rosendale cement mortars, including: the Brooklyn and Washington Bridges in New
York City, the Capitol in Washington, D.C, a multitude of historic forts on the eastern seaboard, and South Carolina cotton mills [9]. The use of other cements or materials for preservation or restoration would not be historically accurate, as well as posing potential problems for the existing materials. Using Rosendale similar to that in initial construction would ensure accuracy, as well as not harm the existing structure.
Distinguishing if one has Rosendale cement in their historical mortar is the first step to deciding to use Rosendale. Methods for analyzing mortar can be found in ASTM C
1324. Chemically a cement mortar can be distinguished from a lime mortar due to the higher silica and aluminum fractions [7]. However, as stated earlier, Portland and
Rosendale cements can be similar in composition. X-ray diffraction[5]or microscopy [7] can be used to determine the presence of alite, thus distinguishing a Portland cement mortar from Rosendale.
8 Strength
Hanley studied the workability of natural hydraulic lime mortars (NHL) and the effects on strength [10]. Three different NHL binders varying in intensity of hydraulic behavior were tested at three different water contents resulting in flows of 165, 195, and 195mm.
Flow is an important property to determine the workability of the mortar, typically a higher flow decreases stiffness and indicates improved workability; however, too high of a flow leads to a watery consistency that is also difficult to work with. Compressive and flexural strengths were calculated at 28-day and 56-day curing periods. The results concluded that stronger binders correlated with stronger mortars, so lime can increase the strength of mortars. Within each category of limes, the one described as “easiest to work with” over “dry” or “runny” also yielded the most desirable strength results. This visual workability test makes making a stronger mortar on-site easy for anyone, but must be kept consistent on a batch-to-batch basis. The difference between “good” and “runny” mortar can vary by less than one percent water content, so consistency is essential.
Papayianni also conducted a study of mortars and their relationships to strength. It is stated that strength is inversely related to porosity in cement mortars by the equation