MNL 67-2009 PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 128
At a glance
Powered by AI
The document discusses the history, origin, types, uses and applications of gypsum.

The document discusses natural gypsum, synthetic gypsum, and insoluble anhydrite as the main types of gypsum.

Gypsum is utilized through processes like dehydration, rehydration and in various applications like plaster, cement, soil conditioning etc.

Gypsum

Connecting Science and Technology


Richard A. Kuntze
Gypsum:
Connecting Science and Technology

Richard A. Kuntze

ASTM International
100 Barr Harbor Drive
PO Box C700
West Conshohocken, PA 19428-2959

Printed in U.S.A.
ASTM Stock Number: MNL67
Library of Congress Cataloging-in-Publication Data
Kuntze, Richard A.
Gypsum: connecting science and technology/Richard A. Kuntze.
p. cm.
‘‘ASTM stock number. MNL 67.’’
ISBN 978-0-8031-7015-5
1. Gypsum. 2. Gypsum—Industrial applications I. Title.
TA455.G9K86 2009
666⬘.92-dc22 2009029218
Copyright © 2009 ASTM International, West Conshohocken, PA. All rights reserved.
This material may not be reproduced or copied, in whole or in part, in any printed,
mechanical, electronic, film, or other distribution and storage media, without the
written consent of the publisher.
Photocopy Rights
Authorization to photocopy item for internal, personal, or
educational classroom use, or the internal, personal, or educational classroom
use of specific clients, is granted by ASTM International „ASTM… provided that
the appropriate fee is paid to ASTM International, 100 Barr Harbor Drive, PO
Box C700, West Conshohocken, PA 19428-2959; Tel: 610-832-9634; online:
http://www.astm.org/copyright.
The Society is not responsible, as a body, for the statements and opinions expressed
in this publication.
ASTM International does not endorse any products represented in this publication.

Printed in Newburyport, MA
October 2009
Contents
1 INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Natural Gypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Synthetic Gypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Insoluble Anhydrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Utilization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4.1 Dehydration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4.2 Rehydration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4.3 Limitations for Utilization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Published Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2 HISTORY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1 Scientific Aspects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Practical Aspects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3 ORIGIN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.1 Natural Gypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2 Natural Insoluble Anhydrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.3 Synthetic Gypsum and Synthetic Insoluble Anhydrite. . . . . . . . . . . . . 18
3.3.1 Flue-Gas Desulfurization Gypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3.2 Phosphogypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.3.3 Other Synthetic Gypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.3.4 Synthetic Fluoroanhydrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

4 BASIC PROPERTIES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.1 Phases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Dihydrate and Dehydration Products. . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.2.1 Dihydrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.2.2 Dehydration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2.3 ␣- Hemihydrate and ␤ -Hemihydrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.2.4 Soluble Anhydrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2.5 Insoluble Anhydrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3 Rehydration to Dihydrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3.1 From Hemihydrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3.2 From Insoluble Anhydrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3.3 Acceleration and Retardation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.4 Misconceptions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.1 Soluble Anhydrite “Prime”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.2 2/3 -Hydrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.3 Colloidal Hydration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.4 Formation of ␣ -Hemihydrate in Stucco. . . . . . . . . . . . . . . . . . . . . . . . . 34
4.4.5 Aging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

5 CALCINATION PROCESSES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.1 Natural Gypsum Preparation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.1.1 Receiving and Storage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.1.2 Drying. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.1.3 Crushing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.1.4 Grinding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.2 Flue Gas Desulfurization 共FGD兲 Gypsum Preparation. . . . . . . . . . . . . . 39
5.2.1 Initial Drying. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

iii
iv Routine Coal and Coke Analysis

5.2.2 Receiving and Storage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40


5.2.3 Final Drying. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.2.4 Grinding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.2.5 Agglomeration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.3 Preparation of Other Gypsum and Anhydrite. . . . . . . . . . . . . . . . . . . . 42
5.3.1 Phosphogypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.3.2 Titanogypsum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.3.3 Insoluble Anhydrite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.4 Calcination Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.4.1 Batch Kettle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.4.2 Continuous Kettles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.4.3 Submerged Combustion Kettles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.4.4 Conical Kettle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.4.5 Rotary Kilns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.4.6 Impact Mill Calciner. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.4.7 Ring Ball and Roller Mills. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4.8 Calcidyne Unit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4.9 Anhydrous and Multiphase Plaster. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4.10 ␣ -Hemihydrate Plasters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

6 PROPERTIES OF STUCCOS AND PLASTERS. . . . . . . . . . . . . . . . . . . . . . . 61


6.1 Stucco Requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6.2 Wet Plaster Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
6.2.1 Normal Consistency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
6.2.2 Setting Characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.2.3 Volume Changes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
6.3 Properties of Hardened Plaster 共Gypsum兲. . . . . . . . . . . . . . . . . . . . . . . . 68
6.3.1 Drying Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.3.2 Apparent Density and Porosity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
6.3.3 Strength. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
6.3.4 Temperature Tolerance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
6.3.5 Fire Protection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

7 GYPSUM BOARD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
7.1 Essential Manufacturing Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
7.1.1 Board Weight. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.1.2 Fire Resistance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.1.3 Board Strength. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
7.1.4 Water Demand. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
7.1.5 Setting Behavior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.2 Core Manufacturing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
7.2.1 Stucco Supply. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
7.2.2 Volumetric Feeder. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
7.2.3 Additive Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
7.2.4 Pin Mixer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
7.2.5 Roll Coater. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7.2.6 Forming Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7.3 Core Composition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
7.3.1 Gauging Water. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
v

7.3.2 Foam System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83


7.3.3 Accelerator System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
7.3.4 Retarders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
7.3.5 Starch. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
7.3.6 Other Additives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
7.3.6.1 Boric Acid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
7.3.6.2 Water Reducing Agents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
7.3.6.3 Paper Pulp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
7.3.6.4 Hydrophobes 共Asphalt Wax Emulsions/Silicones/Siloxanes兲. . . . . . . . 92
7.3.6.5 Glass Fiber, Vermiculite, and Clay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
7.4 Paper Liners. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
7.4.1 Paper Handling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
7.4.2 Paper Liner Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
7.5 Conveyor Section to Dryer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
7.5.1 Board Conveyors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
7.5.2 Knife. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
7.5.3 Acceleration and Transfer Section. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.5.4 Bond between Core and Paper. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.6 Board Drying and Handling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7.6.1 Board Drying. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7.6.2 Dry Board Handling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
7.7 Joint Compounds and Cements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
7.8 Recycling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

8 GLASS MAT GYPSUM BOARD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102


8.1 Sheathing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
8.2 Backing Board. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

9 GYPSUM FIBER BOARD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

10 TESTING, ANALYSIS, AND INSTALLATION. . . . . . . . . . . . . . . . . . . . . . . . 106


10.1 Materials Testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
10.2 Materials Specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
10.3 Product Testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
10.4 Product Specifications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
10.5 Testing Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
10.6 Analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
10.7 Application and Installation Procedures. . . . . . . . . . . . . . . . . . . . . . . . . 111
Preface
The purpose of this manuscript is to close a perceived gap between the aca-
demic and applied aspects of gypsum science and technology since a textbook
covering these subjects is not available. Moreover, academic publications seem
to deal with relatively narrow subjects and are concerned with cataloguing
research results with little applied interpretation. In contrast, the practical lit-
erature is dispersed over a wide range of sources and gives the distinct impres-
sion of being unfocused. It is hoped that closing this gap will help to solve
problems or to clarify situations that technologists and engineers may encoun-
ter and that may not necessarily be well understood.

Therefore, it is the intention to interpret the essential literature on gypsum and


to place the information provided into its proper context. To achieve this, the
characteristic properties of gypsum and cementitious materials derived from it
have been discussed in terms of their strength and weaknesses, taking into
consideration both historical and modern points of view. In any case, this
manuscript is written in a manner which should make it acceptable and acces-
sible not only to a specific audience but also to others generally interested in
this subject.

An equally important reason for preparing this manuscript is the fact that mis-
conceptions continue to resurface on key issues and these are unusually persis-
tent. They touch on all categories from ancient history to modern properties
such as dehydration, ␣-hemihydrate formation, water demand, aging and dis-
integration, rehydration and setting, as well as physical characteristics such as
strength and gypsum board nail pull resistance. This has been the case since
the time of Lavoisier and Le Chatelier in the 18th and 19th century, who first
dealt with the dehydration and rehydration of gypsum in a modern scientific
manner. A contributing factor has been the recent replacement of scientific
research with current issues. Process research has become a most diluted and
misquoted phrase. To a degree, the decline of gypsum research reflects the shift
from gypsum plaster applied in the field to the manufacture of gypsum board in
plants.

vii
Acknowledgements
It is obvious that a publication of this nature cannot be considered without
recognizing the many people who have generously aided me during my involve-
ment with gypsum research since 1955. Otto Schierholtz of the Ontario Re-
search Foundation in Toronto introduced me to the subject of gypsum in gen-
eral and to plaster retarders in particular. Also, Gil Gillespie of Gypsum, Lime
and Alabastine Company in Caledonia, Ontario, as well as Barney Nies and
Jack Summerfield of the United States Gypsum Research Center in Des Plaines,
Illinois, have shared their knowledge over many years while I carried out con-
tract research for their respective companies. I remember with particular grati-
tude that they encouraged publication of scientific papers, which was not com-
mon practice at the time. The members of the ASTM Committee on Gypsum
made available their vast and detailed experience of gypsum and its application
and who permitted me to serve as chairman from 1972 to 1985. I should also
mention Graham Thayer who convinced me to function as chairman of several
conferences on Natural and Synthetic Gypsum in Toronto from 1988 to 2000.
Over most of these periods I have greatly benefited from a professional asso-
ciation with Franz Wirsching of Knauf Gypsum in Iphofen, Germany, with Bob
Bruce of Bob Bruce and Associates, Nanaimo, British Columbia, and Peter
Mayer of CertainTeed Gypsum. Peter Mayer also reviewed the manuscript and
made many suggestions. I am indebted to all of them and many others for their
insight and generosity. Of course, I alone am to blame for any errors and omis-
sions that undoubtedly occurred with the preparation of this manuscript de-
spite every attempt to avoid them.

ix
Terminology
It is fairly well known that the English term gypsum is of Roman origin based
on the Greek term gypsos for which various definitions exist. It is less well
known that gypsum is also defined by archaic terms based on ancient usage, for
example the location from which it is obtained such as alabaster, on descriptive
terms such as the fibrous satin spar, or the specific application for which it is
intended such as selenite and its transparent variety Marienglas. Since gypsum
is a legitimate mineralogical term an attempt is being made in this manuscript
to use it exclusively for the naturally occurring rock as well as synthetic gypsum
and materials obtained from it by grinding, etc.

However, a distinction has to be made between the mineralogical term gypsum


and the chemical term calcium sulfate dihydrate. Both are identical neglecting
the impurities that may be present. Calcium sulfate dihydrate or just dihydrate is
being used whenever purely theoretical concepts are being considered. It is
therefore not uncommon to find the term gypsum used interchangeably with
calcium sulfate dihydrate, even in modern scientific literature. This situation is
being avoided in this manuscript and definitive terms are used wherever pos-
sible. This means that gypsum refers to practical considerations and calcium
sulfate dihydrate to chemical or physical processes.

A similar situation exists with respect to the calcination of gypsum, that is, a
heat treatment to produce stucco 共first calcination stage兲 which consists largely
of calcium sulfate hemihydrate or just hemihydrate. It should be noted that the
term stucco is non-specific since it is also used for other cementitious materials
based on hydrated lime or cement. If the heat treatment involves theoretical
considerations of dihydrate the term dehydration is being used instead of cal-
cination. This means that the term dehydration applies specifically to dihydrate
whereas calcination is used for gypsum.

It should be noted that the term plaster is non-specific as well since it is being
used for other materials such as hydrated lime, cements and a variety of other
materials. However, for this manuscript, the term is used for plaster based on
gypsum. Gypsum plaster is suitable for non-structural interior surfaces. In con-
trast to other materials it is not suitable for structural use or for exterior appli-
cations. Additives such as retarders and stabilizers are added to stucco to form
hardwall plaster, which is mixed with perlite and vermiculite at the factory or
sand and water in the field. In all these cases the term plaster is used, whether
the material is dry or mixed with water. In addition, the term plaster applies to
␣-hemihydrate used for molding and medical/dental applications.

It is also common practice, particularly in Europe, to calcine gypsum com-


pletely to the second calcination stage which means that insoluble anhydrite is
being produced. This dry material is referred to as anhydrite plaster, which is

xi
often blended with stucco. Additives, sand and water are added when this ma-
terial is to be used as finishing plaster for various construction applications.
The term insoluble anhydrite is employed mostly when this chemically active
component is referred to specifically.

More importantly, most of the stucco being produced in North America is pres-
ently used as plaster for the manufacture of paper lined gypsum board 共plas-
terboard兲. For this purpose it is mixed with additives, water and foam. The
term slurry is used for this type of plaster because a larger quantity of mixing
water 共i.e. gauging water兲 is employed than that required to produce a plaster of
normal consistency 共ASTM C472-99兲. Therefore, gauging water is the amount of
water mixed with stucco to produce a plaster with a flow property that is de-
sirable for a specific application.

xii
MNL67-EB/Oct. 2009

1
Introduction
1.1 Natural Gypsum
NATURAL GYPSUM CONSISTING MAINLY OF CALCIUM SULFATE DIHY-
drate is a well-known common non-metallic mineral that occurs as a deposit of
fairly soft and often impure rock. It was deposited from shallow seas as they
evaporated about 200–300 million years ago, mostly during the Permian pe-
riod. This deposition depends on pressure, temperature, and the concentration
of other salts in solution. It is found in most areas near the earth surface and
can be recovered relatively easily using quarries or shallow underground mines.
Natural gypsum is normally white or light gray but depending on the impuri-
ties present can also be pink, dark gray, or almost black. The English term
gypsum is taken from the Roman usage of the Greek “gypsos,” apparently de-
scribing a material that does not burn, although there are other interpretations.
The term may be based on the Arabic “jibs” or “jass,” or the Akkadian 共Meso-
potamian兲 “yasser” or the Persian “gatch.” 共The hemihydrate mineral, called
bassanite is found infrequently.兲
It is common practice to classify gypsum according to its appearance be-
cause it is an ancient material and has been identified by rather archaic terms,
allowing for sub-varieties that are often based on applications. For example, it
can be identified as granular/massive, as satin spar 共fibrous兲 or selenite 共lay-
ered兲. Carving of statues, vases, and other artifacts were often carried out with
a white and nearly translucent variety of fine-grained massive gypsum, called
alabaster probably after one of its sources, Alabastron in Upper Egypt. The
colorless, transparent variety of layered gypsum is cleavable. It can occur in
relatively large plates that were used as window glass. When viewed through it,
the sun gives the appearance of the moon and hence this variety of gypsum was
called selenite. Again there are different interpretations. Similarly, the same
variety of gypsum was used to protect religious images of Mary, originating the
term “Marienglas” 共Fig. 1.1兲. There are also non-commercial ornamental vari-
eties of gypsum known as desert rose, rams horn selenite, etc.

1.2 Synthetic Gypsum


Synthetic gypsum is identical to natural gypsum from a chemical and crystal-
lographical point of view. The difference between the two varieties of gypsum
sources lies primarily in their physical state, which in turn depends on the
origin or the manner of their formation. Synthetic gypsum is generally ob-
tained as the final stage of industrial processes, where sulfuric acid is neutral-
ized by a calcium salt, often CaCO3. Most frequently, however, it is obtained as
flue gas desulfurization 共FGD兲 gypsum during the desulfurization of flue gasses
from coal-fired power plants as FGD gypsum by the reaction of sulfur dioxide
with a calcium salt. Normally calcium carbonate 共CaCO3兲 is used for this pur-

1
Copyright © 2009 by ASTM International www.astm.org
2 Gypsum

Fig. 1.1—Marienglas version of selenite gypsum.

pose or to a lesser extent calcium hydroxide 共Ca共OH兲2兲. In these cases, gypsum


is obtained as a filter cake in the form of wet, fine particles containing usually
about 10 % water.
Therefore, the difference between synthetic gypsum and natural gypsum
lies initially in the original physical condition, which dictates the subsequent
process steps necessary for a particular application. That is, natural gypsum
exists as solid rock, which must be crushed, ground, and dried prior to its
subsequent calcination. In contrast, synthetic gypsum is already finely divided
but it contains more free water than natural gypsum, which must be removed
by drying. Sometimes synthetic gypsum is being agglomerated to facilitate stor-
age, transport, or further processing. It may also be ground after drying, with
or without agglomeration, to change crystal shape and size distribution.
Another significant difference between synthetic gypsum and natural gyp-
sum can be the type and amount of impurities that must be closely monitored
when they occur. For example, pH, chloride content, and carry-over carbon
may be of concern if either FGD gypsum is to be used for gypsum board manu-
facture. The presence of crystalline silica, present in both natural and synthetic
gypsum that may be of concern can best be determined by differential thermal
analysis or differential scanning calorimetry. Phosphoric acid byproduct gyp-
sum has been largely eliminated from further processing because of the pres-
ence of residual phosphates that interfere with the setting of calcined gypsum
共stucco兲. However, the limiting factor for some types of phosphogypsum is the
presence of radium-226.

1.3 Insoluble Anhydrite


It is currently accepted that gypsum, mainly calcium sulfate dihydrate
共CaSO4 · 2H2O兲, is deposited initially and that insoluble anhydrite 共CaSO4兲, the
anhydrous analogue of gypsum, is formed in situ from that gypsum by subse-
quent dehydration. This dehydration involves a rearrangement of the crystal
lattice with complete loss of water that can occur when the temperature rises
due to increases in pressure. However, this subject has always been controver-
sial and numerous theories have been proposed to explain the mixed occur-
Introduction 3

Fig. 1.2—End use of gypsum.

rence of gypsum and anhydrite in layers, lenses, etc. These views are being
complicated since it is possible that natural anhydrite changes back into gyp-
sum by subsequently reacting with groundwater or water seeping in from the
surface. In any case, insoluble anhydrite occurs in the same deposit as gypsum,
sometimes distributed unevenly and in substantial amounts.
Insoluble anhydrite can also be produced by heating gypsum at about
350° C 共662° F兲 and above. This material is identified as thermal insoluble an-
hydrite to differentiate it from the natural variety although both materials are
virtually indistinguishable. However, thermal insoluble anhydrite may exhibit
longer setting times depending to the temperature employed during production
and the length of exposure time. Synthetic insoluble anhydrite is obtained as
fluoroanhydrite in the production of fluoric acid. In at least one situation, this
material is being hydrated and used as gypsum.

1.4 Utilization
Natural gypsum has been used over the ages in an uncalcined form as blocks
for construction purposes similar to ashlars and as the raw material to carve
statues, vases, etc. More recently, utilizations of uncalcined gypsum have in-
volved application as setting time regulator for Portland cement, as fertilizer,
and for soil amelioration. Smaller amounts are used as functional filler for
paints, polymers or paper 共Fig. 1.2兲. These applications are discussed further in
4 Gypsum

Fig. 1.3—Haida Gwaii by Bill Reid. Gypsum plaster pattern for a bronze cast at the
Canadian Embassy in Washington, DC. 共Canadian Museum of Civilization, Ottawa兲.

the video “Gypsum—The Miracle Mineral” 关1-1兴 prepared by the Gypsum Asso-
ciation of Washington, DC.
However, most utilizations of gypsum are based on the fact that it can be
calcined when subjected to heat treatment at fairly low temperatures. This
results in the formation of either stucco as the first calcination stage or in-
soluble anhydrite as the second calcination stage. North American technology
is based almost entirely on stucco. In Europe, both materials are sometimes
produced separately and are then mixed together in certain proportions. When
mixed with water, they rehydrate to gypsum by crystallization, that is, they are
cementitious. In the case of stucco, this takes place in a relatively short time
and in the case of insoluble anhydrite activators are normally used to speed up
the process. Relatively minor quantities of plaster are used as molds for metal
castings 共Fig. 1.3兲 and for dental and surgical applications.
The most modern utilization of gypsum, both natural and synthetic, is for
the manufacture of gypsum board on a large scale 关1-2兴. Gypsum board origi-
nated in the United States, which produces about 45 % of the total amount.
Gypsum board is now produced almost everywhere 共Table 1.1兲. Significantly,

TABLE 1.1—2007 World Capacity for Gypsum Board.


Region/Country Capacity in Million Square Feet Number of Plants
USA 37640 81
Canada 3821 13
Mexico and South America 2270 18
Europe 22250 73
Middle East and Africa 194(?) 11
Australia and New Zealand 2325 13
Japan 7368 23
India and China 8838 29
Total million square feet 84938 246
Total million square meter 7891
Introduction 5

the manufacturing process is based on stucco alone and does not involve in-
soluble anhydrite.
Both dehydration and rehydration phenomena are the basis for much of
the gypsum science and technology.
1.4.1 Dehydration
Under atmospheric conditions, the dehydration of calcium sulfate dihydrate
involves exposure to increasing temperatures with corresponding stepwise loss
of water. It essentially passes through the process indicated below. Dihydrate,
hemihydrate, and insoluble anhydrite are distinct and separate phases since
they have individual crystal structures as illustrated by x-ray deposition
techniques 关1-2兴. Soluble anhydrite, CaSO4 共sol.兲 is a special case but not a
separate phase. It has the same crystal structure as hemihydrate and may be
considered an extension or an anhydrous variation of hemihydrate. Moreover,
soluble anhydrite is unstable and reverts rapidly into hemihydrate when
exposed to water vapor or when coming into contact with liquid water.
Dihydrate→ Hemihydrate→ Soluble Anhydrite→ Insoluble Anhydrite
CaSO4 · 2H2O CaSO4 · 共1/2兲H2O CaSO4 共sol.兲 CaSO4 共insol.兲
Phase I Phase II Anhydrous Phase II Phase III
Stable Metastable Unstable Stable
Depending upon the dehydration conditions, two variations of
hemihydrate are produced, ␣-hemihydrate and ␤-hemihydrate. The
␤-hemihydrate is formed under normal atmospheric conditions independent of
the water vapor pressure present. This is the original variety also known as
stucco in practice. The more recently developed ␣-hemihydrate is formed in an
autoclave or in salt solution between 97° C 共207° F兲 and about 160° C 共500° F兲.
This variety is mainly used for medical or dental applications to prepare molds
as well as in self-leveling floor screeds.
An inordinate amount of work has been carried out on the difference
between these two hemihydrates. In one instance this was done because it was
assumed that the various and changing consistencies 共water demands兲 of
stucco calcined in a kettle were due to different ratios of ␣-hemihydrate to
␤-hemihydrate and that this ratio changed towards ␣-hemihydrate on aging.
The current view is that there is no ␣-hemihydrate formed in the kettle. The
aging of ␤-hemihydrate is due to adsorption of water vapor, which results in a
decrease of the consistency due to the loss of its ability to disintegrate, that is,
to separate into smaller particles when mixed with water. The differences
between the two hemihydrates are now correctly viewed as being mainly due to
the considerably larger size and the uniformity of the crystalline particles of
␣-hemihydrate. This variety of hemihydrate has a low consistency to begin with
and does not disintegrate when mixed with water or age. Consequently its
water demand is not influenced by the adsorption of water vapor.
1.4.2 Rehydration
The rehydration of hemihydrate and of insoluble anhydrite to dihydrate refers
to the solution of these materials in water and subsequent crystallization. This
6 Gypsum

process consists of the following steps with incorporation of water during


crystallization and release of heat.
Hemihydrate + water → Dihydrate + heat 共rapid process兲

CaSO4 · 共1/2兲H2O + H2O → CaSO4 · 2H2O

Insoluble Anhydrite + water → Dihydrate + heat 共slower process兲

CaSO4 共insol兲 + H2O → CaSO4 · 2H2O

1.4.3 Limitations for Utilization


The utilization of gypsum is affected by several overall technical factors, which
play a key role in limiting its application.
1. Structural Effects: Rehydrated gypsum is not a structural material such
as other cementitious materials, e.g., set limestone or Portland cement, because
it is not physically stable. To a large extent, this is due to the high porosity of set
gypsum caused by the evaporation of a larger amount of mixing or gauging
water that is not used for hydration. The gypsum crystals are considered to be
flexible but not elastic, which means that they can be bent but that is difficult or
impossible to straighten them out. Also, some of the contacts between gypsum
crystals are formed by relatively weak hydrogen bonds. As a consequence,
rehydrated gypsum undergoes deformation, such as creep, when subjected to
compressive, tensile, or flexural loads over time. This can be illustrated by the
Instron® “bond”—test where a load is applied to below the breaking point to a
plaster specimen cast in a dumbbell shape. The applied load diminishes over
time, depending on the amount of moisture absorbed by the specimen and the
resulting creep of the specimen. This process can be repeated several times,
which is indicative of a high degree of creep. The original gypsum rock, which
is solid and non-porous, does not show this behavior.
2. Effects of Water: Rehydrated gypsum loses strength in the order of 50 %
when absorbing liquid water, which is much greater than that of other
cementitious materials. It is also sensitive to adsorbed moisture in excess of
true water of crystallization. This is partially due to the higher solubility of
gypsum exacerbated by high porosity and the hydrogen bonding between
gypsum crystals. For this reason, rehydrated gypsum is used primarily for non-
load bearing applications for interior walls and ceilings of buildings. The same
limitation applies when manufactured panels, such as gypsum board or cast
blocks, are employed. However, there are certain advantages associated with
this characteristic. For example, it is possible to bend gypsum boards to a
certain degree, particular with prior wetting, when applying them to curved
surfaces such as pillars. The allowable degree of bending of the wetted board
varies with its thickness whereby thinner boards permit greater bending.
3. Effect of Hydrated Lime: Hydrated lime is often used in combination
with calcined gypsum 共stucco兲 for a variety of reasons. Historically, it was
difficult or impossible to differentiate gypsum from limestone, particularly
since both often occur in mixed deposits. To a degree, this explains the use of
Introduction 7

these materials in applications for which gypsum alone is not suited. At a latter
stage, when the differences were better understood, both materials were mixed
intentionally to improve the properties of gypsum. This tends to hide the
characteristics of gypsum, such as inadequate load bearing properties and
poorer resistance to water. The approach of intentionally using mixtures of
hydrated lime with dehydrated gypsum continued almost to the present day.
4. Effects of Hydration Stage: Insoluble anhydrite rather than
hemihydrate was the product obtained during the early approaches of
manufacture. This is due to the fact that large pieces of gypsum were used for
calcination, which makes it virtually impossible to notice or to stop at the
intermediate hemihydrate level. Also, insoluble anhydrite was actually
preferred since it has a lower water demand and produces longer times to set
and harden, which may have been the intention at an age when set controlling
additives were not available. The conscious use of hemihydrate alone is
dependent upon the calcination of finely ground gypsum, which involves
keeping the material in motion by boiling and stirring. This involved
technology such as the shallow pan kettle that was introduced during the
Baroque or Rococo periods in Europe. The relatively recent application of
hemihydrate as plaster produced in larger kilns or kettles in North America and
to a lesser extent in Europe usually requires the addition of retarders and other
additives to control the setting process.
5. Plaster versus Gypsum Board: A switch from interior gypsum plaster
to gypsum board occurred gradually in North America since the 1950s. This
was less labor intensive and more cost effective. To a large extent it was also
caused by the desire to obtain earlier occupancy of buildings by avoiding the
relatively lengthy drying time required for plasters. In some locations, the
change was also due to the lack of skilled plasterers. However, the key factor
making this change possible was that North America practices a different type
of building method that is based on stud wall construction rather than solid
block or cement walls. Europe, South America, and Asia have lagged initially
behind in gypsum board usage because of their primary residential building
method is based on the latter.

1.5 Published Information


Most recent publications appear to be confined to specific scientific issues and,
consequently, they are dispersed over a wide variety of sources. There seems to
be no recent publication that covers gypsum science and technology in a com-
prehensive manner. Undoubtedly, this is due to the conviction that gypsum is a
well-established building material and that the scientific and technological
points are adequately understood. Needless to say that is not the case. Also,
gypsum research is not considered to be a glamorous field. It is surprising,
however, that the gypsum research that did take place has often itself contrib-
uted to the confusion of specific issues rather than clarifying them. In fact,
certain schools of thought were customarily maintained as if existing in isola-
tion and to the almost total exclusion of information from other research
8 Gypsum

groups. To a large extent, this is also due to the fact that academics in general
do not consult practitioners.
This situation has been summarized as follows in 1941 by Kelly, Southard,
and Anderson: “To explain existing data several varieties of these three types of
compounds have been postulated: two forms of CaSO4 · 2H2O and at least three
varieties of CaSO4 · 共1 / 2兲H2O and four varieties of CaSO4. Moreover, these va-
rieties after being postulated were often confused because of lack of adequate
means of identification. Some varieties were so ill-defined that an unsympa-
thetic critic might claim that a new species of some compound had been pos-
tulated to accompany each set of discordant results.” 关1-3兴. Almost 25 years
later, the situation had changed very little. Kirk and Othmer in 1964 stated the
following: “This substance is probably the subject of more confused and discor-
dant statements, misinterpretations and persistent lack of exact information
than any other compound in the category of inorganic materials.” 关1-4兴.
Against this background, gypsum has been used extensively where it is
locally available. This has occurred despite its natural shortcomings that have
limited its application to interior applications where it does not serve a struc-
tural purpose. From its initial discovery as primitive cementitious material to
its present-day paper lined gypsum board, gypsum technology was advanced by
empirical approaches depending on the particular application. Science has
only played a minor role over the last 200 years.

References
关1-1兴 Gypsum—The Miracle Mineral 共Video兲, Gypsum Association, Washington, D.C.,
1994.
关1-2兴 Weiser, H. B., Milligan, W. O., and Eckholm, W. C., J. Am. Chem. Soc., Vol. 58, 1936,
1261–1265.
关1-3兴 Kelley, K. K., Southard, J. C., and Anderson, C. T., 1941, Technical Paper No. 625,
U.S. Bureau of Mines, Washington, D.C.
关1-4兴 Kirk, R. E., and Othmer, D. S., Encyclopedia of Chemical Technology, 2nd ed., John
Wiley & Sons, New York, 1964, p. 4.
MNL67-EB/Oct. 2009

2
History
2.1 Scientific Aspects
A NUMBER OF EXAMPLES FOR THE UTILIZATION OF PLASTER ARE
reported in the literature from around 7000 to 4000 B.C., such as Catal-Huyuk
in South Central Turkey. However, all these applications are from pre-recorded
history and they are difficult to place into proper perspective primarily because
the available information is non-specific and often contradictory. There is con-
siderable doubt that the plaster used at any of these locations was based on
gypsum alone, but rather mixed with limestone or limestone alone. The differ-
ence between these two materials was poorly understood until Roman times.
A similar picture emerges for the utilization of gypsum in Egypt from 2500
B.C. Limestone and mixtures of limestone and gypsum were converted jointly
and used as floor screed or to repair the exterior of natural stone structures.
This is apparently the case for the pyramid of Cheops, the Sphinx, and subse-
quently the Minoan palace of Knossos on Crete 共2000 B.C.兲.
Gypsum was used in Greece in about 700 B.C. to carve statues and for
various artistic applications. The technology of using the cementitious proper-
ties of gypsum was apparently obtained either from Egypt directly or via Crete.
It is not clear to what extent they understood the difference between limestone

Fig. 2.1—Antoine Lavoisier and his wife and assistant Marie-Anne in a painting by
Jacques-Louis David

9
Copyright © 2009 by ASTM International www.astm.org
10 Gypsum

Fig. 2.2—Le Chatelier

and gypsum as original material or as plaster. They obviously noticed that gyp-
sum was warm to the touch and had a higher solubility as compared to lime-
stone, which was cold to the touch and did not suffer from solubility problems.
Unfortunately, the Greeks failed to leave written records about their uses of
gypsum, which is unusual since they pondered the nature of everything around
them.
As mentioned by Plinius the Elder 共23–79 A.D.兲 关2-1兴 in his “Natural Sci-
ence,” the Romans acquired gypsum technology from the Babylonians and
Greeks, and it was recognized that that there was a difference between lime
plaster and gypsum plaster. The Romans were the first to recognize the limita-
tions of gypsum plaster and used it exclusively for non-structural applications
and for interior purposes such as the interior walls of Pompeii. However, they
differentiated between these two materials by selecting appropriate deposits.
They developed a multilayer gypsum mortar whereby the amount and the par-
ticle size of sand decreased from the bottom to the top layer. This apparently
modern approach tends to reduce shrinkage cracks, a decided advantage when
mortar is used as a base for frescoes or paintings. Unfortunately, the Romans
also left few records about their gypsum technology except for a brief mention
of gypsum by Vitruvius 共80–70 B.C.兲 in his ten books on architecture 关2-2兴.
Gypsum technology was supposedly lost completely with the dissolution of
the Roman Empire in 476 A.D. For this reason, not much is known about the
use of gypsum plaster during the Middle-Ages. For example, mortar based on
History 11

insoluble anhydrite was used in the Harz mountain region during the early
Middle-Ages for the construction of city walls, etc. and to cover interior walls of
buildings. The major construction projects of the 9th to the 13th century were
concerned mainly with monasteries, cathedrals, and palaces. For these gypsum
plastering became more of an art form during the Renaissance period. Towards
the end of the 15th century, towns and cities were built or enlarged. Therefore,
it can be assumed that gypsum technology had been rediscovered at that time.
In any case, it became the special knowledge of a few artisans according to the
custom of the day. Gypsum plastering gained further in popularity during the
Baroque and Rococo periods again mainly for the interior of palaces and
churches. To deal with the increased demand for sophisticated gypsum plaster-
ers, a school was established in Wessobrunn, Bavaria, and a similar school
existed in Italy 关2-3兴. The fire resistive properties of gypsum used for interior
wall covers were recognized early during fires at Basel 共1457兲 and London
共1212, 1666, and 1794兲. King Louis XIV in 1667 decreed that wood structures
had to be covered inside as well as outside with gypsum plaster to resist the
action of fire.
The modern science of gypsum began with the discoveries by Antoine
Lavoisier 共1743–1794兲 outlined in his two papers on gypsum presented to the
French Academy of Sciences in 1765 and 1766. He explained that the heat
treatment of gypsum resulted in the loss of about a fifth of its weight as water
vapor to leave an anhydrous material. He also stated that this material when
mixed with water would form gypsum again by crystallization. It is remarkable
that the reviewers of the Academy praised his work for the “ingenious explica-
tion by which it reduces the phenomenon of the hardening of plasters to the
laws of crystallization” 关2-4兴.
In contrast to his compatriots, Lavoisier employed meticulous experimen-
tation involving weight gain and weight loss of every part of a material when
subjected to heating, which was an established method at the time. He also
combined this process with the novel so-called solvent 共water兲 method. This
was a tremendous advancement primarily because it set aside the concept of
phlogiston 共inflammable兲 proposed by alchemists. They believed that the loss of
weight of a material on heating was due to the irretrievable loss to the air of
phlogiston, an unidentified “component” of that material. Using their preoccu-
pation with metals as an example, this can be expressed by the following for-
mulae: 共1兲 metal 共containing phlogiston兲 → metal calx 共metal oxide兲
+ phlogiston; 共2兲 charcoal 共containing phlogiston兲 → residue+ phlogiston, and
共3兲 metal calx+ charcoal→ metal+ ashy residue. This hypothesis which was
originated by J. J. Becher 共1635–1682兲 and modernized by the German alche-
mist Stahl 共1660–1734兲, was accepted throughout most of the 18th century even
by eminent scientists. Some claim that the phlogiston concept although erro-
neous was the first unifying and comprehensive “scientific” theory whereas
others maintain that it held up scientific development for about 100 years. It is
difficult not to be critical of alchemists but it should be remembered that their
attitudes were by no means uniform and that they were guided by convoluted
concepts such a transmutation of metals, the philosophers’ stone, etc.
12 Gypsum

Lavoisier also helped to set aside Aristotle’s 共384–322 B.C.兲 concept of fire,
air, earth, and water as basic “elements,” as well as the numerous modifications
and opposing views proposed at the time. For this work and other accomplish-
ments as outlined in Méthode de nomenclature chimique and Traité élémentaire,
Lavoisier is rightly known as the Father of Chemistry. As indicated by the quan-
tity of the weight lost when heating gypsum to about 300° C 共572° F兲, Lavoisier
of course obtained insoluble anhydrite. But he observed that 3 / 4 of the water of
crystallization was easier to remove than the last 1 / 4 and that the first 3 / 4 of
the material had a faster hydration than the last 1 / 4 of the material 关2-5兴. As an
extension of Lavoisier’s work, Le Chatelier 共1850–1936兲 described hemihydrate
as the intermediate material and reported that its rehydration to gypsum in-
volves supersaturated solutions with respect to gypsum 关2-6兴. Possibly this is
the reason why hemihydrate and not insoluble anhydrite became identified as
Plaster of Paris. There is no doubt that this was a further tremendous accom-
plishment based on insight and accurate experimentation.
The physical-chemical foundation for the gypsum technology was provided
by J. H. van’t Hoff 共1852–1913兲 关2-7兴 as corrected by Davis 共1907兲 关2-8兴. Consid-
erable confusion had to be subsequently overcome because of the introduction
of the then fashionable colloid theory for the setting process, which was ad-
vanced by quite a number of scientists such as Michaelis 共1893兲, Cavazzi 共1913兲,
Desch and Traube 共1919兲, Ostwald and Wolski 共1920兲, and Baykoff and Neville
共1926兲. However, the work of Weiser et al. in 1936 关2-9兴 showed that colloidal
processes, however defined, were not involved. This meant that water in hemi-
hydrate occupied specific locations in a crystal lattice. This also meant that
hemihydrate is not a zeolite, which would have water occupying spaces in a
random fashion. The need for clarification concerning hemihydrate and soluble
anhydrite resulted in the work of Kelly, Southard, and Anderson in 1941 关2-10兴.
Unfortunately, they followed suggestions by Riddell, which led to a misinter-
pretation of the origin and role of ␣-hemihydrate and ␤-hemihydrate 关2-11兴.
It is difficult to establish when and where stucco based on hemihydrate
alone was prepared and used separately from insoluble anhydrite. However, it
certainly occurred during the 19th century. The common procedure of calcin-
ing gypsum was to place fairly large pieces of gypsum on a grate with an open
fire underneath. This produces primarily insoluble anhydrite because it makes
exact temperature control rather difficult and takes no account of the water
vapor released. In contrast, stucco was produced at the same time by finely
grinding gypsum and heating it in shallow pans such as the Harz mountain
kettle. Under these conditions, the water vapor released has a two stage fluid-
izing effect. The end of the first effect can be used to determine the end of the
calcination process to hemihydrate. However, it should be realized that there is
some overlap of the phases so that stucco usually contains a few percent of
dihydrate and soluble anhydrite.

2.2 Practical Aspects


Stucco based primarily on hemihydrate was prepared in the above manner and
introduced as a modern plastering material. There are many cases in Europe
History 13

where this material has been used for the decoration of interior of walls and
ceilings. However, for the normal coverage of walls the plastering industry in
Europe continued to preferred mixtures of hemihydrate and anhydrite such as
the building plaster 共Surcuit兲 in France and a similar product 共Putzgips兲 in
Germany. There are several reasons for the preference of these mixtures over
hemihydrate plasters. The most important one seems to be that less water is
used to prepare these plasters and their strength is correspondingly higher.
Also, a desirable delay of the hardening process is achieved without or with
small amount of retarding additives, such as tartaric acid and hydrated lime. In
some instances, hemihydrate plasters were employed as mortars, mixed with
various quantities of sand or lightweight aggregates. Stucco based on hemihy-
drate was used exclusively for the manufacture of blocks and tiles.
The development in North America was quite different. To a large extent,
this can be explained by the difference in construction practices, which are
based on stud wall construction in North America versus solid brick or block
wall construction in Europe. Lime plaster was used initially to finish interior
walls and ceilings. This material hardens slowly but it requires a wood lath for
mechanical keying. Waste lumber from saw mills cut into strips and nailed to
wood framing became the standard plaster base. However, around 1885, it was
discovered that retarders based on soluble keratin could be used to control the
setting of hemihydrate and to make gypsum plaster with a quicker hardening
process than lime plasters. The ability to retard plaster based on stucco was the
breakthrough required for its widespread application. For example, gypsum
plaster in North America, and to some extent in Britain, consisted of a two or
three coat application containing sand, with a final finishing coat of neat 共un-
sanded兲 lime/gypsum plaster instead of the European one coat application. 共Ex-
cept for Keene’s cement, anhydrite plaster is unknown in North America.兲
Because of the many shortcomings of wood lath it was replaced by a gyp-
sum lath patented by Augustine Sackett in 1894 关2-12兴. This lath or Sackett
board consisted of four plies of paper and three layers of stucco plaster. Despite
some problems with this lath, plants were built in several places in the United
States. In 1909 these plants were purchased by U.S. Gypsum who developed
Sackett board further by introducing a variety of new developments such as the
paper folded edge. Subsequently, an Adamant board was developed consisting
of a gypsum core faced with top and bottom paper. To meet increasing con-
struction demands, further changes were made in 1926, which adopted the
round edge and introduced foam and fiber into the gypsum core. This product
has become known as Rocklath and it exists today in various adaptations.
The conversion from lath to gypsum board started in the early 1950s. It was
driven primarily by two issues: 共1兲 Avoiding the drying time of plaster which
allowed earlier occupancy of buildings and 共2兲 the lack of skilled plasterers in
many locations. Gypsum board in its many manifestations has now replaced
plaster almost entirely except for special applications. About 1 % of the cal-
cined gypsum produced in the United States is used to manufacture plaster
while nearly 26 million tons were used in 2005 to produce 36 000 billion square
feet of gypsum board. The consumption of gypsum board has been increasing
14 Gypsum

Fig. 2.3—Augustine Sacket

substantially since in 2000 only 21 million tons of calcined gypsum was used to
produce 26 000 billion square feet of board. North American gypsum board
systems have also been accepted in Europe, Asia, and South America particu-
larly for high rise and institutional construction.

References
关2-1兴 Plinius Secundus, G., Naturalis Historia 36, 24 共77兲.
关2-2兴 Vitruvius, P., De Architectura Libri Decem 共10 Books on Architecture兲, Akademie-
Verlag, Berlin, 1964.
关2-3兴 Stark, J. and Wicht, B., History of Building Materials, Bauverlag, Wiesbaden and
Berlin, 1998.
关2-4兴 Bell, M. S., Lavoisier in the Year One, W. W. Norton, New York and London, 2005.
关2-5兴 Lavoisier, A. L., “The Analysis of Gypsum,” Royal Acad. Sci., Vol. 1765, pp. 76–81.
关2-6兴 Le Chatelier, M. H., C. R. Hebd. Seances Acad. Sci., Vol. 96, 1883, pp. 1668–1671.
关2-7兴 van’t Hoff, J. H., Armstrong, E. F., Hinrichsen, W., Weigert, F., and Just, G., “Gyp-
sum and Anhydrite,” Ztschr. Physikal. Chem., Vol. 45, 1903, pp. 257–306.
关2-8兴 Davis, W. A., “The Nature of the Changes Involved in the Production and Setting of
Plaster of Paris,” J. Soc. Chem. Ind., London, Vol. XXVI, No. 13, 1907, pp. 727–738.
History 15

关2-9兴 Weiser, H. B., Milligan, W. O., and Eckholm, W. C., J. Am. Chem. Soc., Vol. 58, 1936,
pp. 1261–1265.
关2-10兴 Kelley, K. K. and Anderson, J. C., 1941, Technical Paper 625, U.S. Bureau of Mines.
关2-11兴 Riddell, W. S., Rock Products, May 1950.
关2-12兴 Sackett, A., “Inside-Wall Covering,” U.S. Patent No. 520,123 共1894兲.
MNL67-EB/Oct. 2009

3
Origin
3.1 Natural Gypsum
AS MENTIONED BEFORE, GYPSUM WAS DEPOSITED ABOUT 200–300
million years ago during the evaporation of seawater that was probably repeat-
edly cut-off from the main body. This crystallization process is influenced by
many factors but essentially it takes place in brine of various concentrations
and not in plain seawater. In Europe gypsum deposits formed primarily during
the Permian, Triassic, and Jurassic periods. They occur in France, Spain, Ger-
many, and Italy whereas the one near Paris is best known because of the name
Plaster of Paris. In North America gypsum was deposited during the Ordovi-
cian and Silurian periods as well. The best known deposits in Africa are the
Adrigat formation in the Blue Nile Valley, which dates from the Jurassic period,
and the deposit of the Ethiopian plateau 共Danakil depression兲 关3-1兴. Smaller
deposits of gypsum are abundantly present almost everywhere except in areas
of volcanic origin. Gypsum formation takes place by crystallization following
the relative solubilities of the other salts present. Accordingly, gypsum is depos-
ited after calcium carbonate and prior to sodium, magnesium, or potassium
chlorides and other sulfates. Natural gypsum deposits often contain primary
impurities in various quantities, such as calcium sulfate anhydrite, calcium
carbonate, dolomite, or various types of clay. Minor amounts of primary impu-
rities that are found less frequently are silica, bitumen, glauberite
共Na2Ca共SO4兲2兲, and syngenite 共CaSO4 · K2SO4 · H2O兲. Secondary impurities are
introduced after deposition by materials in solution seeping into cavities or by
contamination during the mining process.
Gypsum usually forms a solid nonporous rock near the Earth’s surface and
consequently the mining of gypsum ore is generally carried out in quarries or
shallow underground mines. Gypsum can be pure white, gray, or reddish if clay
is present or if it contains iron oxide, or nearly black if bitumen is present.
Based on differences in appearance or structure that reflect crystallization pat-
terns and crystal size, gypsum is also known by a variety of sometimes archaic
terms, such as massive including alabaster 共fine grained/translucent兲, selenite
共sheet-like/transparent兲, or satin spar 共fibrous兲. Because of its white color, it is
also known as terra alba or as landplaster because of its use for agricultural
purposes. Earthy gypsum or gypsite forms as a secondary deposit when gyp-
sum is remixed with other material after the original deposition. Gypsum sand
is usually contaminated with Glauber’s salt, Na2SO4 · 10H2O, or Glauberite.
Both physicochemical and geological factors have to be taken into consid-
eration when trying to explain the nature and configuration of the gypsum
deposits, in particular those that are stratified repeatedly with insoluble anhy-
drite. Physicochemical factors are the result of short term laboratory exercises
that study primarily the effects of various salt concentrations and temperature.

16
Copyright © 2009 by ASTM International www.astm.org
Origin 17

Geological factors on the other hand try to take into account long term geologi-
cal events as well. These involve periodic flooding of seas that may be only
partly or temporarily isolated, subsiding basins, and tectonic lifting that oc-
curred repeatedly over millions of years. Not surprisingly, the two factors dis-
agree in many respects.
However, none of the salt concentrations and ratios studied in the labora-
tory is found under evaporite conditions in the field. Accordingly, there has
been no success in reproducing real geological situations with laboratory ex-
periments. For example, if one assumes an average salinity of sea water, only
about 1.5 % of the original volume is deposited on evaporation. This means
that for a depth of 1000 m, only 15 m of salt is deposited. Of these 15 m only
0.55 m are gypsum. This contrasts with deposits that can be as much as 600 m
thick.
Dihydrate forms double salts with alkali-metal and ammonium sulfate,
some of which occur in nature, such as syngenite, CaSO4 · K2SO4 · H2O. An
example for a triple salts is the polyhalite, CaSO4 · MgSO4 · K2SO4 · 2H2O,
which is also found in nature. A variety of other triple salts exist that involve
divalent ions of iron, zinc, manganese, copper, and magnesium, as well as alkali
metal and ammonium.

3.2 Natural Insoluble Anhydrite


The problem of deposition is highlighted when considering the occurrence of
insoluble calcium sulfate anhydrite within gypsum deposits. This is a complex
proposition involving salt concentration and temperature as well as other fac-
tors. Anhydrite is found as layers with various thicknesses on top, at the bottom
or between layers of gypsum. It is also found intermingled and as lenses within
gypsum deposits apparently without indication of differential precipitation. Ac-
cording to physicochemical consideration the precipitation of gypsum or anhy-
drite is a matter of the CaSO4 concentration depending on the temperature
intervals and taking into account the stabilities of the two phases. For example,
during sea water evaporation at 30 ° C 共86 ° F兲, gypsum will always precipitate
when the salinity lies between 3.35 and 4.48 times the normal salinity of the sea
water. Anhydrite is deposited above this concentration. On the basis of this and
other considerations, it is also postulated that anhydrite will be deposited
above 42 ° C 共108 ° F兲. Unfortunately it is doubtful that this concept can be used
to explain the occurrence of anhydrite in gypsum deposits since it is improb-
able that 42 ° C 共108 ° F兲 was reached and exceeded over the required long
periods of time.
The Blaine formation in Oklahoma was studied microscopically and petro-
graphically and it represents an opposite view. According to the results, the
gypsum deposit in the area was formed near the surface by the hydration of
anhydrite. It follows that the anhydrite was the initial precipitate in an evapo-
rating basin. Unfortunately there is no other supporting evidence available. For
example, no consideration is given to the volume increase when anhydrite is
changed into gypsum, which would stop it at a certain depth.
An interesting case is the formation of giant crystals 共up to 11 m兲 of gyp-
18 Gypsum

sum in a limestone cave at the Naica mine near Chihuahua, Mexico. These
gypsum crystals were apparently formed from anhydrite during long term ex-
posure to 58 ° C 关3-2兴.
However, considering the available data and opinions, it becomes obvious
that there is no easy answer to the question of whether gypsum or anhydrite
was deposited first. Nevertheless, the current view is that gypsum was the origi-
nal material deposited and that anhydrite was formed from it at a later time.

3.3 Synthetic Gypsum and Synthetic Insoluble Anhydrite


Synthetic gypsum is being produced by a variety of industrial processes. Most
of these obtain synthetic gypsum by a neutralizing reaction between sulfate and
calcium ions such as in the manufacture of phosphoric acid or titanium diox-
ide. Most processes that use sulphuric acid to digest rocks must be neutralized
before they can be used or safely discarded. Limestone is the best case of a
neutralizing agent. As mentioned previously, synthetic gypsum is chemically
identical to natural gypsum and has the same crystal structure. Although it has
usually a higher dihydrate content 共purity兲 than most natural gypsums, its im-
purities are different and that can be of concern 关3-3兴.
Despite the availability of synthetic gypsum, there has been an initial reluc-
tance in North America to use this material in the gypsum industry. These
concerns are often based on costs for transportation and drying. However, the
key reason for this reluctance was the question as to whether these materials
would perform as well as natural gypsum, particularly for the manufacture of
gypsum board. There was also a question of sufficient supply in appropriate
locations. Nevertheless, over the past 20 years, there has been a considerable
change toward the use of synthetic gypsums that were obtained by the desulfu-
rization of flue gases at coal fired power plants. Because of the slow start in the
North America to use these materials, the industry in Japan and in Europe was
initially more advanced. 共FGD gypsum is known as Desulpho-Gypsum in Brit-
ain and as REA-Gips in Germany.兲 For example, in 1983 Japan used 2.5 million
tons per year of FGD gypsum whereas the total combined consumption in
Europe and North America was only 0.3 million tons per year. In 2005 this has
risen in North America to over 9 million tons per year or 34 % out of a con-
sumption for gypsum board or plaster manufacturing of 26 million tons per
year. With the exception of the relatively small amount of titanogypsum and
flouroanhydrite, other synthetic gypsums or synthetic anhydrites are still not
being used.
A number of circumstances have contributed to make the acceptance of
some synthetic gypsum a reality in North America. The greatest impetus was
provided in the United States by the Clean Air Act of 1963, its 1970 Extension,
as well as similar provisions. This led to the production of considerable quan-
tities of acceptable quality synthetic gypsum primarily from coal fired power
plants using the limestone or lime process. The other main reasons were the
consolidation of the gypsum industry, the tendency toward the building of
larger plants for the production of gypsum board rather than plaster. In some
Origin 19

Fig. 3.1—Consumption of total gypsum and synthetic gypsum in the United States.

cases, the depletion of suitable natural gypsum supply played an important role
共Fig. 3.1兲.

3.3.1 Flue-Gas Desulfurization Gypsum


The largest quantities of so-called FGD gypsum are presently obtained from the
desulfurization of flue gases from power stations and other large production
plants that use fossil fuels such as bituminous coal. Almost all of these operate
by the counter current washing of flue gas with an aqueous suspension of
limestone or lime to remove the sulfur dioxide 共Fig. 3.2兲. Numerous other
processes have been developed for this purpose but only a few have been
industrialized.
For limestone scrubbing the net reaction is represented by

SO2共g兲 + CaCO3共s兲 + 共1/2兲O2共g兲 + H2O → CaSO4 · 2H2O共s兲 + CO2共g兲 .

The first step takes place at a pH of 7–8 and produces an almost insoluble
calcium sulfite

Fig. 3.2—Desulfurization of the gases in a coal fired power plant.


20 Gypsum

共1兲 SO2共g兲 + CaCO3共s兲 + 共1/2兲H2O → CaSO3 · H2O共s兲 + CO2共g兲 .


In the second step the suspended calcium sulfite is converted at pH 5 to soluble
calcium bisulfite

共2兲 2CaSO3 · H2O共s兲 + 2SO3共g兲 + H2O → 2Ca共HSO3兲2 共solution兲 .


In the third step the calcium bisulphate is oxidized at pH 5 to calcium sulfate
dihydrate, the so-called FGD gypsum

共3兲 Ca共HSO2兲2 共solution兲 + H2O → CaSO4 · 2H2O共s兲 + H2SO4 .


The remaining sulfuric acid produced reacts with the remaining limestone,
thus forming additional gypsum

共4兲 H2SO4 + CaCO3共s兲 + H2O → CaSO4 · 2H2O + CO2共g兲 .


Large and uniform gypsum crystals normally form in the aqueous
suspension. These crystals or crystal clusters are separated from the aqueous
suspensions by dewatering in hydrocyclones and vacuum belt filters or by
centrifuges. This produces a filter cake of gypsum having a free water content
of slightly less than 10 %. Characteristically, FGD gypsum has a high purity
共high percentage of calcium sulfate dihydrate兲 and an off-white color. The
fineness and the crystal shape of the gypsum are important factors for the
manufacture of gypsum board and they may require corrective steps. As always
the quality of the gypsum is of concern. Problems in that respect are generally
centered on the pH of the material, its chloride content, and in some case the
flyash and residual carbon content. About 4.5 tons of gypsum is produced per
ton of sulfur in the fuel. Apart from small quantities used in Portland cement,
all of the FGD gypsum currently produced is consumed by the gypsum industry
primarily for the manufacture of gypsum board 共Fig. 3.2兲.
3.3.2 Phosphogypsum
Worldwide about 130 million tons per year of phosphogypsum are currently
produced. This constitutes the largest quantity of any type of synthetic gypsum
and it is expected to increase as the demand for fertilizers increases. In any
case, the quantity of phosphogypsum produced per year exceeds by far the
demand on gypsum and anhydrite, which is only about 75 million tons
annually.
Usually phosphogypsum is produced during the production of wet
phosphoric acid from phosphate rock 共apatite兲 and sulfuric acid: Ca5共PO4兲3F
+ 5H2SO4 + 10H2O → 5CaSO4 · 2H2O + 3H3PO4 + HF.
Normally about 1.7 tons of dihydrate are produced per ton of raw
phosphate. This corresponds to 5 ton of dihydrate per ton of phosphorus
pentoxide produced.
The phosphogypsum is obtained as a moist fine powder with a free water
content of 20 %–30 %. Considerable amounts of impurities are usually present
depending on the source of the rock and the particular process used. Of major
concern is the radioactivity of the phosphogypsum as produced, which again
depends to some extent on the rock source. For this reason Florida rock is
Origin 21

much less suitable for utilization than rock from Morocco or Kola. For
example, during the decay of radium-226 into radon-222 and its other decay
products, an increase in the dose rate occurs in the inside of buildings. For this
reason, the utilization of phosphogypsum from Florida is banned since it
contains approximately six times the limits set by EPA.
In North America, technical difficulties in removing radioactive materials
are the main reason why currently only about 4 % of the phosphogypsum
produced is utilized mostly in agriculture or as road base. No phosphogypsum
is used in any construction material such as gypsum board or Portland cement.
It is an extremely expensive and involved process to remove unwanted
substances from phosphogypsum. The other principal difficulty preventing the
use of phosphogypsum is its high content of free water and subsequently high
drying costs.
Another serious problem that is often overlooked is that isomorphic
incorporation of chemical constituents into the dihydrate lattice can occur
during the wet phosphoric acid process. Most commonly calcium hydrogen
phosphate, CaHPO4 · 2H2O, is incorporated because it has almost the same
lattice spacing as dihydrate. It is also possible to incorporate monosodium
phosphate, Na2HPO4, or fluorophosphates, FPO3, into the dihydrate lattice.
Hemihydrates produced from phosphogypsum of this type are characterized by
abnormal set times and poor strength development. This is the main reason for
the failure of the Giulini process for ␣-hemihydrate. The process did not
remove sufficient phosphate to improve the properties of the hemihydrate even
with recrystallization in solution.

3.3.3 Other Synthetic Gypsum


Small quantities of calcium sulfate are formed during the production of
organic or inorganic acids, such as tartaric acid, citric acid, oxalic acid, boric
acid, or titanium dioxide. Larger quantities of synthetic gypsum are expected to
be produced because of the treatment of sulfate-containing waste waters and
the neutralization of dilute sulfuric acids that occur during the processing of
minerals such as zinc ore. However, for various reasons these gypsums are
difficult to utilize. The exception is the relatively small amount of synthetic
gypsum which is obtained during the manufacture of titanium dioxide
pigment. This material is characterized by a small particle size, high purity, and
a white color.
Attempts continue to be made to utilize gypsum scrap from the ceramic
industry and metal foundries. Again, the largest problems are created by the
removal of impurities and water introduced during slip casting. Although
gypsum processing is known to be forgiving, it cannot tolerate certain soluble
impurities or impurities in large quantities.

3.3.4 Synthetic Fluoroanhydrite


This is the only representative of insoluble anhydrite which is being
manufactured by reacting fluorspar with sulfuric acid during the production of
hydrofluoric acid. This process results in the formation of fluoroanhydrite
according to the following reaction:
22 Gypsum

CaF2 + H2SO4 → CaSO4 + 2HF.


1 ton of fluorspar produces about 1.75 tons of anhydrite. The total
production of this material is about 2–3 million tons per year. Fluoroanhydrite
is being utilized only in Europe primarily for self-leveling floor screeds. If
aluminum sulfate is used as accelerator, hydration to gypsum is often difficult,
which to a large extent is due to isomorphism that may occur between ions of
aluminum fluoride 共AlF−− −−
5 兲 and sulfate 共SO4 兲.

References
关3-1兴 Antolini, P., Proceedings of RILEM Symposium on Calcium Sulfates and Derived Ma-
terials, Saint-Remy-Les-Chevreuse, France, May 25–27, 1977.
关3-2兴 Garcia-Ruiz, J. M., Villasuso, R., Ayora, C., Canals, A., and Otálora, F., Geology, Vol.
35, 2007, pp. 327–330.
关3-3兴 “Analyses of FGD Gypsum,” VGB-M-701, Technical Society of Large Power Plant
Operators, 1991 共in German兲.
MNL67-EB/Oct. 2009

4
Basic Properties
4.1 Phases
THE CaSO4 – H2O SYSTEM IS REPRESENTED BY THREE SOLID PHAS-
es that exist at room temperature in air containing water vapor, namely cal-
cium sulfate dihydrate 共CaSO4 · 2H2O兲, calcium sulfate hemihydrate
共CaSO4 · 共1 / 2兲H2O兲, and insoluble anhydrite 共CaSO4-insol.兲. In these cases a
phase is identified and characterized by an individual and distinct crystal struc-
ture. Soluble anhydrite 共CaSO4-sol.兲 is not a phase because it has the same
crystal structure as hemihydrate. In fact, it is considered to be the anhydrous
state of hemihydrate. However, because of its theoretical and industrial impor-
tance, it is discussed together with the above three phases.
Two underlying factors are always evident when discussing these sub-
stances and in particular the changes from one into the other. They are 共1兲
dehydration of the dihydrate and 共2兲 rehydration to the dihydrate. Dehydration
refers to the stepwise loss of water of hydration that is accomplished by in-
creasing the temperature of dihydrate for the purpose of obtaining one of its
dehydration products, i.e., hemihydrate, soluble anhydrite, and insoluble anhy-
drite. The temperature ranges given for the individual dehydration steps de-
pend on whether the process is static or kinetic. In general, static conditions
appear to proceed at a lower temperature but they require more time. For
example, dihydrate begins to dehydrate at about 46° C 共115° F兲 but conversion
takes months or even years to complete.
Rehydration is the exposure of these dehydrated materials 共hemihydrate,
soluble anhydrite, and insoluble anhydrite兲 to liquid water for the purpose of
obtaining the dihydrate again at standard conditions. This process proceeds by
the crystallization of the dihydrate in water. If the starting material is hemihy-
drate, a supersaturated solution with respect to dihydrate is formed prior to
crystallization. In the case of insoluble anhydrites as the starting material, the
influence of activators and dehydration temperature shows that the rate of
conversion to gypsum is affected by these parameters in addition to solubility.
The solubilities of dihydrate, hemihydrate, and insoluble anhydrite are im-
portant consideration for a variety of reasons but primarily because they allow
a distinct separation between the three phases on a theoretical basis. Their
approximate solubility curves are shown in Fig. 4.1. The intersection between
two curves gives the equilibrium temperature below which the less soluble
phase is stable. For example, the curves show that above 40° C 共104° F兲 共point
A兲 insoluble anhydrite is the most stable phase. Above this point, dihydrate
when in contact with water does not convert to insoluble anhydrite. In practice,
it may do so but at a very slow rate. Below 100° C 共212° F兲 共point B兲 dihydrate
is more stable than the hemihydrate. Therefore, saturated solutions of hemihy-

23
Copyright © 2009 by ASTM International www.astm.org
24 Gypsum

Fig. 4.1—Solubility of calcium sulphates.

drate are supersaturated with respect to dihydrate, and dihydrate precipitates


from that solution.
This dehydration-rehydration process appears to be relatively simple but
there are a number of significant circumstances to be considered. For example,
water vapor plays a significant role in the dehydration and rehydration of
soluble anhydrite 共Fig. 4.2兲. Also, hemihydrate occurs in the ␣-form and ␤-form
depending on the method of dehydration 共Fig. 4.3兲. Insoluble anhydrite can be
produced by dehydration but it also occurs as the natural material. High tem-
perature anhydrite exists only above 1180° C 共2166° F兲 and is not discussed
further because it is not of interest for the purposes of this manuscript.
Exposure of hemihydrate to water vapor in the atmosphere results in the
adsorption of the water to a maximum of slightly less than 2 % of the material
depending on its surface area. Dihydrate is not formed under these conditions
unless the amount of adsorbed water vapor exceeds about 2 %, which leads to
capillary condensation. This nonstoichiometric water can be completely re-
moved by drying below 46° C 共115° F兲.
Basic Properties 25

Fig. 4.2—Effect of water vapor pressure on soluble anhydrite formation.

4.2 Dihydrate and Dehydration Products


4.2.1 Dihydrate
Dihydrate is the basic phase in the CaSO4 – H2O system. Its crystal structure
reflects a ratio of one molecule of calcium sulfate to two molecules of water,
hence its chemical name. It occurs as monocline crystals with four molecules
per unit cell: a = 5.67 Å, b = 15.15 Å, c = 6.51 Å, ␤ = 118° 23⬘. The unit cell
structure is a lattice layer, in which calcium ions 共Ca2+兲 and tetrahedral sulfate
ions 共SO42−兲 form strong double sheet that remain essentially unchanged
during phase changes. These sheets are completely separated by double layers
of water of crystallization. This means that the water is embedded in planes
between the calcium sulfate layers which allows for easy cleavage. It also
means that the water of crystallization is essential for the integrity of the entire
structure and to some extent explains the ease with which dihydrate is
converted to hemihydrate.
The water constitutes about 21 % of the total mass of the dihydrate but it
represents almost 50 % of its volume. This condition also contributes to the
26 Gypsum

Fig. 4.3—Differential thermograms of ␣- and ␤-hemihydrates.

ease with which dihydrate is converted to hemihydrate by heating 共Table 4.1兲.


Flörke provided the most modern version of the crystal structure of dihydrate,
hemihydrate, and insoluble anhydrite, including corrections of earlier work
关4-1兴 关Fig. 4.4兴.
The thermodynamical stabilities of dihydrate and its dehydration products
are classified as follows: dihydrate= stable, hemihydrate= metastable, soluble
anhydrite= unstable, and insoluble anhydrite= stable. This classification refers

TABLE 4.1—The CaSO4 – H2O system.


Hemihydrate Soluble Insoluble
Characteristics Dihydrate ␣-hemihydrate ␤-hemihydrate Anhydrite Anhydrite
Formula CaSO4 · 2H2O CaSO4 · 共1 / 2兲H2O CaSO4 (sol.) CaSO4 (insol.)
Molecular mass 172.17 145.15 136.14 136.14
Thermodynamic Stable Metastable Unstable stable
stability ⬍46° C ⬎46– 1180° C
Water of 20.92 6.21 0 0
crystallization
Density, g / cm3 2.31 2.76 2.62–2.64 2.58 2.9–3.0
Hardness, Mohs 1.5 3–4
Solubility in 2.1 6.5 ⬇8.8 Hydrates to 2.7
water at 20° C, hemihydrate
g/L
Lattice Monoclinic Rhombohedral (Deleted) Orthorhombic
symmetry
Refractive index 1.522
Basic Properties 27

Fig. 4.4—Crystal structure of dihydrate, hemihydrate, and insoluble anhydrite.

to conditions at room temperature in air that contains water vapor. Under these
conditions, only the dihydrate is stable below 46° C 共115° F兲. Above this
temperature only the insoluble anhydrite is stable. Hemihydrate is metastable
since it changes relatively rapidly into dihydrate in the presence of liquid water.
Soluble anhydrite is unstable because in the presence of water vapor or liquid
water it converts to hemihydrate almost instantaneously.
The solubilities of dihydrate and the various dehydration products have a
major effect on the characteristics of these substances. For example, solubility
determines the general behavior of dihydrate in comparison to other
cementitious materials that are much less soluble in water. Specifically, the
high water solubility of dihydrate makes it unsuitable for external applications
or any situation where exposure to water is a possibility. In addition, the
solubility difference between dihydrate 共 ⬃ 2 g / L兲 and that of hemihydrate
共 ⬃ 8 g / L兲 is the major reason for the crystallization of the latter at room
temperature. Insoluble anhydrite has a lower solubility than dihydrate above
about 46° C 共115° F兲, which is reflected in its hydration behavior.
The solubility of dihydrate is influenced greatly by acids, especially sulfuric
acid, phosphoric acid, and nitric acid. Dihydrate is readily soluble in glycerol
and in aqueous solutions of chelating agents, such as EDTA. The solubilities of
the other forms of calcium sulfate are strongly affected by the presence of other
compounds.
A larger number of papers have been published that deal with the reaction
of acids, salts, etc., and the formation of compounds with calcium sulfate. This
work is not covered here because it is largely of academic interest.
4.2.2 Dehydration
Dehydration is the first of the two key features of dihydrate science and
technology. When carried out under normal atmospheric conditions,
hemihydrate, soluble anhydrite, and insoluble anhydrite are produced from the
28 Gypsum

TABLE 4.2—Heats of dehydration.


Phase change from dihydrate J/mol kJ/g
To ␤-hemihydrate 86,700 597,200
To ␣-hemihydrate 84,600 582,700
To ␤-anhydrite (sol.) 121,800 895,700
To ␣-anhydrite (sol.) 117,400 863,100
To anhydrite (insol.) 108,600 798,000

dihydrate at progressively increasing temperatures. Examples of the formation


of the phases and states are given below. The temperature ranges are based on a
variety of practical considerations and are approximate

Dihydrate → hemihydrate → soluble anhydrite → insoluble anhydrite


100 – 180°C → 180 – 350°C → 350 – 1180°C
The dehydration mechanism shown reflects a kinetic or dynamic process. It
does not take into account long exposure times or slow rates of conversion. For
example, as a result of the high water content and the sandwich structure,
dehydration of dihydrate occurs at fairly low temperatures. The lowest
temperature at which dehydration can proceed is 46° C 共115° F兲, but it does so
at a very slow rate. Also, insoluble anhydrite is formed from the other phases
above 46° C 共115° F兲 but its conversion requires a very long time.
These processes can be described as follows:

Dihydrate → Hemihydrate → Soluble Anhydrite → Insoluble Anhydrite


(requires liquid water, (converts with water (requires liquid water,
rehydrates relatively fast vapor or liquid water to rehydrates relatively slow
converts into dihydrate) hemihydrate) converts into dihydrate)

The dehydration kinetics of the dihydrate substantiates these reaction


processes. Kelly et al. made a thorough study of the thermodynamic properties
of the CaSO4 – H2O system. Table 4.2 gives the heats of dehydration of the
various phase changes that are of significance. However, all the hypothesis
concerning reaction mechanisms, activation energies, and orders of reaction
have not yet been proved. The problems are that the kinetics of the phase
changes is often inhibited and that the test conditions have not been precisely
defined.

4.2.3 ␣- Hemihydrate and ␤-Hemihydrate


When the temperature of dihydrate is raised, the first phase obtained is
hemihydrate. As the name implies, it has a ratio of one molecule of calcium
sulfate to 0.5 molecule of water. During dehydration the volume is decreased
and channels are created that run parallel to the CaSO4 chains. As a result, this
dehydration step involves a change in crystal structure from a sandwich to a
honeycomb format. This honeycomb structure is unique in that water
molecules are not essential for its integrity. That is, the water molecules occupy
Basic Properties 29

TABLE 4.3—Technological properties of ␣- and ␤-hemihydrates (ASTM C26 or DIN 1168).


Consistency 共mL/ 100 g兲 35 (alpha) 42 (alpha) 66 (beta) 90 (beta)
Linear expansion on setting (m/m) 0.0028 0.004 0.0021 0.0016
Setting time (min) 15–20 15–20 20–40 25–35
Flexural strength—dry 共kg/ cm2兲 [psi] 125 [1778] 49–50 [697–711]
Tensile strength dry 共kg/ cm2兲 [psi] 66 [931] 13 [185]
Tensile strenght—1 h—wet 共kg/ cm2兲 [psi] 35 [499] 6.6 [94]
Compressive strength—dry 共kg/ cm2兲 [psi] 560 [7965] 450 [6400] 100–130 [1422–1849] 56 [797]
Compressive strength—1 h—wet 共kg/ cm2兲 [psi] 280 [3982] 28 [398]

specific spaces stoichiometrically but they do so without being strongly


bonded. This means that hemihydrate is not a zeolite, which is characterized by
the absence of a stoichiometric arrangement.
When hemihydrate is obtained under normal atmospheric conditions, it is
referred to as ␤-hemihydrate to differentiate it from ␣-hemihydrate. The latter
is a form obtained from dihydrate either by autoclaving or by solution
processes. An inordinate amount of work has been carried out to establish the
differences between these two forms using most available instrumental
methods.
In the past, this has led to the proposition that both forms represent two
limiting states whereby intermediate states can be prepared by changing the
water vapor pressure existing during dehydration. From a practical point of
view this work was done because ␣-hemihydrate has a considerable lower
water demand than ␤-hemihydrate when preparing a plaster with a standard
consistency. It was believed, therefore, that the different consistencies of kettle
calcined stucco could be explained by different ratios of ␣- to ␤-hemihydrate or
alternatively by a species that lies between the two extremes. Moreover, it was
erroneously believed that aging, that is, the irreversible lowering of the
consistency of kettle calcined stucco when exposed to adsorbed water, could be
explained by a change of ␤-hemihydrate into ␣-hemihydrate.
Of the techniques used to identify the difference between ␣- and
␤-hemihydrate 共heat of hydration, stacking order index, DTA, consistency兲, only
a few still remain valid. The chief difference seems to be due to the fact that the
␣-hemihydrate particles consist of relatively large uniform mono-crystals. In
contrast, the ␤-hemihydrate particles consist of clusters of micro-crystals that
have the ability to undergo aging. That means that in the presence of adsorbed
water vapor they lose the ability to disintegrate 共change into individual micro-
crystals兲 which gives the appearance of a change towards ␣-hemihydrate.
The data given in Table 4.3 show that at the first approximation the
compressive strength of ␣-hemihydrate and ␤-hemihydrate depends largely on
the amount of water used to prepare specimens, such as cubes, and not on the
modification. The data also show that the compressive strengths of wet
specimens are about a 50 % lower than that of dry specimens. This is a much
greater effect than obtained for other cementitious materials such as Portland
cement.
30 Gypsum

TABLE 4.4—Heats of hydration.


Phase Change to Dihydratte J/mol J/g
From ␤-hemihydrate 19,300± 85 111.9± 0.50
From ␣-hemihydrate 17,200± 85 100.0± 0.50
From ␤-anhydrite (sol.) 30,200± 85 173.3± 0.50
From ␣-anhydrite (sol.) 25,700± 85 149.6± 0.50
From anhydrite (insol.) 16,900± 85 98.0± 0.50

Differential thermal analysis clearly shows the difference between the two
forms 共Fig. 4.3兲. The thermograms indicate that the conversion of
␣-hemihydrate to insoluble anhydrite takes place immediately following the
dihydrate to hemihydrate conversion rather than at 350° C 共662° F兲 关4-2兴.
However, this difference may be lost when grinding the ␣-hemihydrate.
Frequently quoted differences in heats of hydration are given in Table 4.4.

4.2.4 Soluble Anhydrite


Soluble anhydrite is obtained by removing the water molecules from the
honeycomb structure of the hemihydrate, usually on further heating. This
occurs without changing the original crystal structure. However, this process is
dependent on the water vapor pressure present. That is, for each temperature
up to about 200° C 共392° F兲, a threshold water vapor pressure exists above
which soluble anhydrite does not form. However, soluble anhydrite does form
almost instantaneously when the temperature exceeds this threshold water
vapor pressure. This is due to the unique characteristic of the hemihydrate,
which allows water molecules to leave and re-enter its structure with relative
ease despite the fact that they occupy specific stoichiometric positions. There
appears to be no difference between ␣- and ␤-hemihydrate in this respect.
However, above about 200° C 共392° F兲 even saturated water vapor does not
prevent soluble anhydrite formation. Below about 350° C 共662° F兲 the
dehydration-rehydration process is reversible and repeatable so that soluble
anhydrite or hemihydrate can be obtained whenever either the temperature or
the water vapor pressure is changed 关4-3兴.
Since the soluble anhydrite is simply the anhydrous state of hemihydrate,
the cumbersome term “dehydrated hemihydrate” is sometimes used to describe
it. Care must be taken during its preparation because it reacts rapidly with
water vapor at room temperature and even at elevated temperatures up to
about 200° C to form hemihydrate. Also, insoluble anhydrite begins to form
above 100° C 共212° F兲 within a relatively short period of time. Therefore,
soluble anhydrite must be prepared within certain limits of time, temperature
and water vapor pressure. Significantly, soluble anhydrite converts rapidly to
hemihydrate when exposed to water vapor or to liquid water. This
characteristic is utilized industrially and in the laboratory when Drierite®
共soluble anhydrite兲 is used to remove water from liquids and gasses. Spent
Drierite® 共hemihydrate兲 can be restored by heating at appropriate
temperatures.
Basic Properties 31

4.2.5 Insoluble Anhydrite


The third phase, insoluble anhydrite, is obtained by raising the temperature of
hemihydrate or soluble anhydrite beyond about 350° C 共662° F兲. As mentioned
before, it must also be kept in mind that insoluble anhydrite can form slowly at
lower temperatures. This dehydration step involves a further change in crystal
structure from the honeycomb format of ␤-hemihydrate into a compact format
of the insoluble anhydrite. Consequently, its structure lacks channels and for
this reason it is the least reactive phase of the calcium sulfates. It exhibits the
closest packing of ions, which makes it the densest and strongest of the calcium
sulfates. The crystal structure becomes more compact with increasing
temperature and time of exposure, which results in increasingly longer times of
reacting with water.
Insoluble anhydrite produced by heat treatment is often called thermal
insoluble anhydrite 共or “dead burn” in industry兲 to differentiate it from natural
insoluble anhydrite. Both varieties of insoluble anhydrite have the same crystal
structure. Also, both varieties are similar in appearance and are being used in
similar applications.

4.3 Rehydration to Dihydrate


Rehydration is the second key feature, which refers to the addition of liquid
water to the hemihydrate or insoluble anhydrite. As mentioned before, soluble
anhydrite converts instantaneously to hemihydrate and behaves thereafter as
hemihydrate.
4.3.1 From Hemihydrate
For the rehydration process, hemihydrate dissolves to form a supersaturated
solution with respect to dihydrate. It crystallizes as dihydrate from this solution
and solidifies when the appropriate amount of water is used, producing only a
slight expansion. The solidification appears to involve three types of bonds: 共1兲
true crystalline intergrowth where ions are shared, 共2兲 network formation by
growing past each other into water filled spaces, and 共3兲 hydrogen bonding
where crystals touch. The time of conversion of hemihydrate to dihydrate is
dependent on the calcination method but is usually in the order of 20 min. This
time can be lengthened to several hours by the addition of retarders or
shortened to less than 3 min by using suitable accelerators. The utilization of
hemihydrate plasters depends to a large extent on these two characteristics.
That is, in the past hemihydrate plastering technology required the use of
retarders and stabilizers 共retarder/accelerator combinations兲 whereas the
present technology of manufacturing gypsum board relies mostly on the use of
accelerators to achieve the required short setting times.
4.3.2 From Insoluble Anhydrite
Insoluble anhydrite crystallizes directly into dihydrate without forming a
solution that is significantly supersaturated with respect to dihydrate. An
increase in volume of about 60 % has been estimated for this process. This
means that the process is essentially governed by the rate of solution of
insoluble anhydrite rather than its solubility. Conversion normally takes place
32 Gypsum

in a matter of days. This time can also be shortened to hours by the addition of
activators such as K2SO4 but it is difficult to obtain setting times in the range
of minutes. Obviously, this confines the use of insoluble anhydrite to certain
applications such as mining mortars.

4.3.3 Acceleration and Retardation


The setting times of hemihydrate and of insoluble anhydrite are shortened or
lengthened by the addition of accelerators and retarders. Both effects are
absolutely essential for their application or to the manufacture of products,
such as hardwall plaster and gypsum board. Accelerators reduce the time
required for hemihydrate to form dihydrate whether or not retarder has been
used previously to lengthen the setting time. This demonstrates again that
gypsum is resilient and forgiving under certain circumstances.
The most common accelerator is gypsum itself by providing homogeneous
nucleation points. This effect is noticed most often in the field when plaster is
contaminated with freshly set gypsum. 共Heterogeneous nucleation occurs when
acceleration takes place with a material other than gypsum.兲 Ground gypsum
treated with sugar, starch, or a surfactant is used for the manufacture of
gypsum board. So is potassium sulfate, which represents a chemical type of
accelerator by initially providing a large amount of sulfate ions. Aluminum
sulfate in the presence of lime is used to accelerate and stabilize the set of
hardwall plaster in the presence of retarder 关Fig. 6.22共4兲兴. A common retarder
in the past was soluble keratin, which was obtained by hydrolyzing animal hoof
or horn material by treatment with sodium and calcium hydroxide. This
process provides relatively long chain molecules with carboxyl, sulfo, amino,
and amide groups. However, during a study of the mechanism of retardation
关4-4兴, it was shown that short molecules are quite effective if the functional
groups are in a preferred position. Carboxyl groups are always absolutely
essential as shown by octric acid 共1, 1⬘, 3, 5, 5⬘-pentane pentaacetic acid兲,
which is an extremely potent retarder. For the manufacture of gypsum board,
soluble keratin has been largely replaced by diethylene triamine pentaacetic
acid 共DTPA兲 关4-5兴. This type of retarder represents a series of synthetic
materials, which have demonstrated the function of amino groups if carboxyl
groups are in the correct position. DTPA can be fed directly into the gauging
water just ahead of the pin mixer and results in better control of retarder
addition. The above retarders are characterized by a delay of the onset of the
hydration with a minor effect on the rate of hydration itself. Another type of
retarder performs differently in that it delays the rate of hydration with a minor
effect on its onset. This type of retarder is represented by natural tartrate,
which is activated by the presence of calcium hydroxide. Interestingly, it can be
used with insoluble anhydrite either alone or mixed with stucco. In addition,
several chemical species which represent small molecules are being used to
retard the rate of hydration of plaster. This includes some phosphates which
must be handled with a certain degree of care because of their potentially
negative effect on the strength of plaster.
Basic Properties 33

4.4 Misconceptions
4.4.1 Soluble Anhydrite “Prime”
Various materials have been characterized by x-ray analysis, differential
thermal analysis, scanning electron microscopy, mercury porosimetry, and
measurement of the specific surface area. On the basis of the values obtained,
the existence three limiting states for soluble anhydrite were postulated:
␤-anhydrite, ␤-anhydrite prime, and ␣-anhydrite. An ␣-anhydrite prime was
considered as well. In particular, it was claimed that ␤-anhydrite prime is
produced at very low water vapor pressures when the water vapor is released
rapidly and the particle size is small. Under these conditions, ␤-soluble
anhydrite “prime” forms directly from the dihydrate without formation of an
intermediate hemihydrate. Not surprisingly the ␤-anhydrite “prime” was
shown to have a much higher specific surface area than either the ␤-anhydrite
or ␣-anhydrite 关4-6兴.
However, it should be obvious that soluble anhydrites and hemihydrates
can be prepared in the laboratory with vastly different properties and
characteristics. Consequently, this work merely identifies one method of
preparing a ␤-soluble anhydrite. It must be emphasized that it does not
constitute a separate phase but simply another condition since both
hemihydrate and soluble anhydrite have the same crystal structure. This is a
classic example of how confusion can be created about the existence of the
various phases, states or forms in the CaSO4 – H2O system. Moreover, it would
be necessary to use a different nomenclature for hemihydrates obtained from
the soluble anhydrite prime by absorption of water vapor. For example, if this
material differs in their physical properties from conventional hemihydrate it
should be designated as ␤-hemihydrate prime.

4.4.2 2 / 3-Hydrate
Attempts are being made repeatedly to claim the existence of a calcium sulfate
2 / 3-hydrate. The last proposal was made in the late 1990s by studying the
structure of dehydration products 关4-7兴. Specifically, it was claimed that both
hemihydrate and the 2 / 3-hydrate were obtained during the dehydration of
dihydrate and the crystallographic particulars of the 2 / 3-hydrate were
identified. However, on closer examination of the data presented, it turns out
that only the possibility of the existence of 2 / 3-hydrate was originally claimed,
not the existence of the material itself 关4-8兴. Therefore, its existence cannot be
accepted without considering additional evidence.

4.4.3 Colloidal Hydration


Cavazzi 关4-9兴 and several others proposed a colloid theory according to which
the hydration of hemihydrate to dihydrate proceeds via an intermediate
colloidal or non-crystalline state. This concept is based primarily on the
observation that hemihydrate plaster passes through a “plastic” state during the
very early hydration stage where no dihydrate is yet detectable. The colloid
theory has also been modified to include a topochemical reaction, which
34 Gypsum

Fig. 4.5—Stacking order indices of ␣- and ␤-hemihydrate.

includes a hydration of the hemihydrate particles taking place over very short
distances.
There are several other explanations for the behavior of the hemihydrate
plaster, such as dispersion or disintegration of particles. The most likely
explanation, however, is that that microcrystalline substances with a particle
size smaller than 1 ␮m can behave as a colloid. This is due to the fact that
below that particle size the surface of the substance determines its properties
instead of its mass.

4.4.4 Formation of ␣-Hemihydrate in Stucco


To explain the differences in the water demand and setting time of kettle
stucco, it has been postulated that ␣-hemihydrate and ␤-hemihydrate are
formed in stucco during calcination. In principle, it was postulated that the
ratios of these two forms were dependent upon the water vapor pressure
surrounding the particles during the calcination process 关4-10兴. Heat of
hydration measurements were used to support this hypothesis 关4-11兴. An
inordinate amount of work was carried out subsequently to study the effect of
water vapor on the formation of ␣-hemihydrate or ␤-hemihydrate. This led to
various other hypotheses including the existence of a range of hemihydrates
exhibiting intermediate properties between the two forms. The most interesting
concept was advanced by Morris 关4-12兴 who found differences in stacking order
indices 共Fig. 4.5兲.
However, none of these hypotheses have stood the test of time.
Consequently, it is currently believed that only ␤-hemihydrate is formed during
kettle calcination or at any atmospheric condition. The differences in
properties between various stuccos produced and the changes with time are
Basic Properties 35

more properly assigned to differences in the reactivities of the stucco particles,


as exhibited by the degree of dispersion, aging, etc.

4.4.5 Aging
Aging of stucco is a process that over time causes the drastic reduction in water
demand 共consistency兲 and the increase of setting time due to the absorption of
water vapor 关4-13兴. The reduction of water demand is due to the loss of ability
of stucco particles to disintegrate seemingly forming larger particles as the
aging process proceeds 关4-14兴. Initially, this process has been assumed to be due
to the conversion of soluble anhydrite to hemihydrate, which occurs very
quickly in the presence of absorbed water vapor. However, it has been
demonstrated that two separate processes occur and that aging of stucco
proceeds long after all of the soluble anhydrite has been converted.
It was further assumed, following a suggestion by Riddell 关4-10兴, that the
loss of water demand on aging involves the conversion of ␤-hemihydrate into
␣-hemihydrate. The foundation for these assumptions was again provided by
Kelley et al. based on heat of hydration measurements. However, it was shown
by Lane’s work that exposure to absorbed water vapor leads to a reduction in
water demand, which is due to the loss of stucco particles that can be
disintegrated. The increase in setting time over the same period is caused by
the same changes of the apparent reactivity of the stucco crystal aggregates.
Dihydrate is formed only when the amount of adsorbed water exceeds 2 %. This
is a somewhat contentious issue because dihydrate can form after long periods
of time 共days and weeks兲 at moisture levels below 2 %.
A newer version of aging has been introduced through the work of Bold
关4-15兴, in which stucco is exposed for a short period of time 共minutes兲 to steam
to lower its consistency.

REFERENCES
关4-1兴 Flörke, O. W., “Crystallographic and X-Ray Studies in the System CaSO4·2H2O,”
Neues Jahrbuch, Abh., Vol. 84, 1952, pp. 189–240.
关4-2兴 Powell, D. A., “The Transformation of the ␣- and ␤-Form of Calcium Sulphate
Hemihydrate to Insoluble Anhdrite,” Nature 共London兲, Vol. 182, 1958, pp. 792–793.
关4-3兴 Kuntze, R. A., “Effect of Water Vapor on the Formation of CaSO4·共1/2兲H2O Modifi-
cations,” Can. J. Chem., Vol. 43, 1965, pp. 2522–2529.
关4-4兴 Kuntze, R. A., “Retardation of the Crystallization of Calcium Sulphate Dihydrate,”
Nature 共London兲, Vol. 211, 1966, pp. 406–407.
关4-5兴 Kuntze, R. A., “Synthetic Plaster Set Retarder,” U.S. Patent 3,304,189 共Februay 14,
1967兲.
关4-6兴 Lehmann, H. and Rieke, K., “The System CaSO4-H2O,” Tonind. Ztg., Keram. Rund-
schau, Vol. 97, 1973, pp. 157–159.
关4-7兴 Hummel, H.-U., Abdussaljmanow, B., Fischer, H.-B., and Stark, J., “The Stability of
Calcium Sulfate,” Proceedings of the 14th International Conference on Building Ma-
terials, 2000, 1-0067, Bauhaus University Weimar.
关4-8兴 Pedersen, B. F. and Semmingsen, D., “The Gypsum-Anhydrite Equilibrium,” Acta
Crystallogr., Section B: Struct. Crystallogr. Cryst. Chem., Vol. 38, 1982, pp. 1047–
1077.
36 Gypsum

关4-9兴 Cavazzi, A., “The Calcium Sulfate Relationships,” Z. Chem. Ind. Kolloide, Vol. 12,
1913, pp. 196–201.
关4-10兴 Riddell, W. C., Rock Products, May 1950.
关4-11兴 Kelley, K. K., Southard, J. C., and Anderson, C. T., U.S. Bureau of Mines Technical
Paper 625, 1941.
关4-12兴 R. J. Morris, “Stacking Order Indices of ␣- and ␤-Hemihydrate,” Nature 共London兲,
Vol. 198, 1963, pp. 1298–1299.
关4-13兴 Kuntze R. A., Materials Research and Standards, ASTM, West conshohocken, PA,
1976, Vol. 7, p. 350.
关4-14兴 Lane, M. K., Rock Products, March 1968.
关4-15兴 Bold J. “Method and Apparatus for Stabilizing Plaster,” U.S. Patent No. 7,371,278
B2 共May 13, 2008兲.
MNL67-EB/Oct. 2009

5
Calcination Processes
GYPSUM PROCESSING CAN BE CONSIDERED TO ENCOMPASS TWO
separate stages. The first consists of the preparation of the raw gypsum and
covers such steps as receiving and storage, drying, crushing, and grinding. The
second stage involves the calcination of this material using a variety of equip-
ment, such as kettles and kilns. Obviously, the essential feature of the first stage
is to prepare the raw gypsum so that it is acceptable by the calcination unit.
There are many procedures to accomplish this but it is always necessary to
control moisture content and particle size distribution. For this reason, the
procedures discussed below should be considered as examples to accomplish
these tasks.

5.1 Natural Gypsum Preparation


Natural gypsum is mined in quarries using drilling and blasting at several levels
or by underground mining using chamber blasting. This rock consists of large
pieces usually containing up to 4 % free moisture. The coarse rock is conveyed
to the crushing plant often located at the gypsum plant site. Impact crushers,
jaw crushers, and single roll crushers with screens and oversize returns are all
suitable for coarse size reduction. Impact pulverizers or roller mills are used for
intermediate size reduction, and hammermills, ball mills, and ring-roll mills
are used for fine grinding. The mined and crushed gypsum rock is sometimes
homogenized before being calcined. This is done in homogenization plants,
with capacities of about 1 week production. Cleaning or beneficiation of gyp-
sum is done infrequently at the present time. A possible exception is the re-
moval of Glauber’s salt 共Na2SO4 · 10H2O兲 by washing.
5.1.1 Receiving and Storage
Conventional gypsum processing is considered to begin when the raw material
is received by the plant. For example, 35,000 ton self-unloading ships, 100 ton
rail cars, or 25 ton trucks, generally supply gypsum rock with a maximum
diameter up to about 6 in. Conveyors generally move rock delivered by ship to
a storage site. Because natural rock is relatively coarse, rain and snow does not
materially increase its moisture content. This material has usually a density of
90 lb/ ft3 共1442 kg/ m3兲 and an angle of repose of 38°, which is given by the
angle between the base of the material and its slope. This angle can be
increased to 45° if the rock is dried and its particle size is reduced. Outdoor
stockpiles of 1,200,000 ton are not unusual while storage buildings usually have
a maximum capacity of about 40 000 ton. The intention is to provide a supply
of 14 days for the plant. If transport to distant gypsum or cement plants is
necessary, the gypsum is reduced to the size required by those plants.
Rock supplied by ship is often transferred by conveyor belt from the
storage site to drying and crushing section of the plant. Rock supplied by rail

37
Copyright © 2009 by ASTM International www.astm.org
38 Gypsum

car or truck may be stockpiled or fed through a grizzly into a hopper to be


conveyed to the plant. Other feeders, such as apron conveyors or vibrating
pans, may be used to assure an even feed to the belt conveyor.

5.1.2 Drying
If the free moisture content of the gypsum rock is below 3 % it can usually
bypass the dryer and move directly to the crushing stage. However, for rock
with moisture content above 3 % some drying may be necessary. Rock obtained
from outdoor storage often contains a higher moisture content, which must be
removed or reduced prior to crushing. The drying of rock can be obtained with
a directly heated rotary dryer operating at 180° F 共82° C兲. This type of dryer is
usually 6 ft 共1.8 m兲 in diameter by 50 ft 共15 m兲 long and is capable of reducing
the moisture content of the rock from 8 %–10 % to 2 %–3 % at a rate of
50 ton/ h 共45 metric ton/ h兲.
5.1.3 Crushing
Natural rock is usually reduced in size to about 3 / 8 to 1 1 / 2 in.
共0.95– 3.81 cm兲. In many plants, the rock passes through a screen and only
material larger than 1 in. 共2.54 cm兲 is crushed. This approach allows the use of
a lower capacity mill because only about one-third of the rock is larger than
1 in. in size. Both fractions are then combined and transported by bucket or
screw conveyors to a crushed rock storage bin.
Excessive amounts of dust can be produced by high-speed impact crushers
and by overloading of crushers. Crushers should therefore be operated at the
lowest possible speed. For this reason, conveyors are normally enclosed for
dust control and hoods leading to baghouses are located at each transfer point.
Electrostatic precipitators or cyclones are used in some plants but they are
considered inferior to baghouses. Magnets may be provided before and after
crushing to remove ferrous materials 共nails, wires, etc.兲. The bucket elevators or
screw conveyors generally supply the crushed rock to several storage bins.
Vibrators or air blast systems maintain proper flow. The bins typically have a
capacity for a 24-hour operation of the plant. In some plants, facilities are
provided to allow intermediate storage of the crushed rock. This provides more
flexibility if other uses are contemplated. For example, the crushed rock can be
homogenized if necessary at this stage. Also, the rock can be screened and the
appropriate fraction used in rotary kilns or flash calcining mills. Further, it is
possible to supply rock for Portland cement from these sources.

5.1.4 Grinding
Crushed rock is fed into a roller mill system, which is designed to reduce the
rock to an appropriate size range. As a general guide, 90 %–95 % passing a
100-mesh 共150 ␮m兲 screen is normally obtained. A variety of other mills can be
used for this purpose as well. At this stage, it is possible to further reduce the
moisture content of the rock by drying. In a more specific example, dried rock
is passed through a Raymond roller mill to produce “landplaster” with a
fineness of 0-trace on a 40-mesh screen 共0.425 mm兲 and 60 %–65 % through a
325-mesh 共0.045 mm兲 screen. Retention on a 100-mesh screen is then
Calcination Processes 39

established to meet the 325-mesh criteria. Both 40-mesh and 100-mesh values
are then measured frequently. The use of a 50-mesh screen is common as well.
The ground gypsum is often called landplaster because of its traditional use
as fertilizer and soil conditioner. It is collected in cyclones and fed by screw
conveyors to the kettle feed bins. These bins should contain enough material to
fill the calcination kettle several times. High-speed hammer mills and single roll
crushers normally produce higher proportions of fines than slow-speed
hammer mills, ring roll mills or gyratory, jaw or double roll crushers.

5.2 Flue Gas Desulfurization „FGD… Gypsum Preparation


In North America quite a number of companies are now using synthetic gyp-
sum as a complete replacement of natural gypsum. In 2006, about 46 % of the
total consumption of 35 million tons of gypsum consisted of synthetic mate-
rial. In almost all cases, this involves FGD gypsum. Titanogypsum is used in
only a few locations in combination with other types of gypsum to compensate
for the small particle size of this material. Large quantities of FGD gypsum are
now available at power plants that are often located near population centers.
One reason for accepting synthetic gypsum is its high purity 共dihydrate con-
tent兲 combined with a low level of deleterious impurities. This makes these
materials suitable for the manufacture of gypsum boards. Also, FGD gypsum
can be processed into conventional stucco, multiphase plasters as well as
␣-hemihydrate plasters. In some cases, the plaster may require additional
preparation or blending to obtain the required plasticity as a slurry.
In general, the potential for substitution of natural gypsum by FGD gypsum
in a specific location depends on many factors including gypsum quantity and
quality, proximity to power plants, lack of disposal sites, disposal costs, envi-
ronmental concerns, and government regulations. In addition, the relationship
of the user of the gypsum to the utilities is of major importance. However, after
overcoming some initial problems, it seems that FGD gypsum has now found
general acceptance.
FGD and natural gypsum differ in the chemical composition of their impu-
rities. They also have different physical properties that may require special
attention. In general, FGD gypsum has a higher purity and a lower level of
major impurities such as limestone or dolomite. However, it may have a higher
content of potentially deleterious impurities such as sodium, magnesium, and
chloride salts as well as calcium sulfite, CaSO3 · 共1 / 2兲H2O, flyash and residual
carbon. It may also have a higher pH than 8, which can produce a poor bond of
core to paper. It is important that impurities are below a specified level.
FGD gypsum requires a somewhat different handling procedure because of
a number of differences between the two materials. The key difference is the
fine particle size and narrower particle size distribution of the FGD gypsum. It
also has a moisture content that is higher, usually just below 10 % as compared
to natural gypsum about 4 %.
5.2.1 Initial Drying
Primary dewatering is the separation of the solid gypsum from the scrubbing
slurry, which is the first step of preparing FGD gypsum for subsequent use.
40 Gypsum

Gypsum crystals obtained in a forced oxidation system are relatively large,


which allows hydrocyclones to be used rather than thickeners that are less
efficient. The hydrocyclones normally concentrate the solids from the slurry at
the scrubber outlet from about 10 % to about 40 %.
Horizontal vacuum filters or vertical basket centrifuges are used for
secondary dewatering by the majority of the recently constructed forced
oxidation FGD plants. A cost analysis for this type of equipment and its
operation should include the cost for chloride removal by washing and
subsequent treatment of the wash water. Different types of dewatering
equipment require different quantities of wash water and different auxiliary
equipment for washing the filter cake.
The free water content of the filter cake from a horizontal vacuum belt filter
is typically 10 %–12 %. A basket batch centrifuge produces free water content
of about 6 %–8 %. FGD gypsum treated by both types of dewatering systems
can be handled by dump trucks. Dust problems may arise when the moisture
content falls below 5 %.

5.2.2 Receiving and Storage


FGD gypsum requires a covered storage because the amount of moisture held
or adsorbed is a function of the particle size. Consequently, rain or snow may
increase the moisture content of FGD, which will make subsequent processing
difficult and more costly.
FGD gypsum is generally supplied as a wet filter cake or alternatively as a
dried and agglomerated product or reclaimed from wet or dry storage stocks. It
is seldom received as a completely dry powder or wet agglomerate. If the
material is supplied as a wet filter cake, additional facilities will be required for
unloading and subsequent handling. However, extensive modification of
equipment and facilities is required if the material is supplied as a dry powder.
Material that is dried and agglomerated requires the same equipment as
natural rock because of the similarity in size.

5.2.3 Final Drying


In general, drying of FGD gypsum is carried out prior to most calcination
processes. The rotary dryer employed for drying natural gypsum rock can
possibly be used in some cases. However, it may require the same modification
as carried out for the cage mill or the Williams mill because of the higher water
content and the fine particle size of the material. Therefore other drying
equipment such as the fluidized bed dryer has been used. This type of
equipment can also be used for the combined drying and calcining.
The fluidized bed dryer is similar to the Hazemag rapid dryer or the
Babcock flash dryer used in Europe. It consists essentially of a closed vessel
with two rotating shafts in the lower part. The FGD gypsum is fed into the
dryer behind the hot gas inlet. The hot gases entering the vessel has a
temperature of about 400° C 共752° F兲 and exits at about 90° C 共194° F兲. The
gypsum is moved upwards and forward by the rotating paddles with the hot
gases to the end of the dryer. The residence time of the gypsum ranges from
8 to 13 s. The dried gypsum is then calcined in kettles and the stucco produced
Calcination Processes 41

can be used without additional grinding for the manufacture of gypsum board
or blocks. However, the particle size of the gypsum is too small to produce
stucco to be combined with anhydrite to form plaster. One of the directly
heated, co-current dryers for non-agglomerated FGD gypsum is shown in Fig.
5.1 as part of a drying-agglomeration plant.

5.2.4 Grinding
The grinding of FGD gypsum may be necessary if modification of its crystal
shape or its particle size distribution is required. This improves the handling
characteristics of the gypsum or the properties of the stucco after calcination.
For example, changes in the size distribution of the gypsum may be required to
change its fluidization and flow properties, which are important properties
during the calcination process. This can possibly be achieved by increasing the
⬍10 ␮m proportion of the gypsum, either by grinding or addition of fine
gypsum or limestone. Also, the shape and size of the FGD gypsum crystals can
be permanently modified by briquetting at 100 kN/ mm2 and subsequent
grinding. This suggests that briquetting is a suitable process to prepare FGD
gypsum with poor crystal shape for conventional calcination.
The small particle size of the FGD gypsum makes it particularly suitable for
kettle calcination. It may also be suitable for special rotary kilns that use
ground gypsum, but these are not common. One such rotary kiln,
manufactured by Schilde 共now part of Grenzebach BSH兲, is used in North
America. Other exceptions are indirectly heated tube kilns also provided by
Babcock-BSH.

5.2.5 Agglomeration
For conventional calcination in the rotary kiln or the conveyor kiln, FGD
gypsum must be agglomerated before it is suitable as feed material. Various
agglomeration processes have been developed for FGD gypsum that primarily
supply, with subsequent grinding, larger particles for specific calcining
processes. Similarly, certain Portland cement plants can be provided with this
material.
Of the agglomeration methods such as granulating, extrusion, pelletizing,
and briquetting, the latter has proved to be the most effective and is the most
widely accepted. In this process, the FGD gypsum must be dried prior to
compacting without the use of binders or other additives. This produces a 2 cm
thick 共3 / 4 in.兲 and 6 cm long 共2 1 / 2 in.兲 briquette. Grinding of the briquettes
also provides the means of modifying the crystal shape and size distribution of
the FGD gypsum. However, this involves a high cost of about $10/ ton. Dried
and agglomerated FGD gypsum can be handled by conventional equipment in
the same way as natural gypsum rock; assuming that fine particles created by
excessive attrition do not create unforeseen difficulties.
The briquettes have mechanical properties that do not differ substantially
from those of natural gypsum. They have a hardness of 500 N, an apparent
density of ⬇2.15 g / cm3 and a bulk density of about 1 ton/ m3 共68.6 lb/ ft3兲.
Also, they are exceptionally abrasion resistant and they can be stored in the
open since they do not disintegrate when exposed to rain or frost. For this
42 Gypsum

Fig. 5.1—FGD gypsum drying and briquetting: 共a兲 hot as inlet; 共b兲 moist FGD gypsum
inlet; 共c兲 rapid dryer; 共d兲 hot gas outlet; 共e兲 cyclone; 共f兲 dry FGD gypsum inlet; 共g兲
compacting press; 共h兲 briquettes.

reason, FGD briquettes can be processed in conveyor kilns to multiphase


plasters in the same manner as natural gypsum. They can be used alone or
mixed with natural gypsum.
Figure 5.1 shows a flow diagram of a plant for processing FGD gypsum by
drying and briquetting. The extra energy required to dry one tonne of dry FGD
gypsum consists of 550,000 kJ of thermal energy and 12 kWh of electrical
energy. Agglomeration by briquetting requires 10 kWh.

5.3 Preparation of Other Gypsum and Anhydrite


5.3.1 Phosphogypsum
The utilization of phosphogypsum is problematic mainly because it contains a
high proportion of deleterious impurities such as phosphates, silicofluorides,
and organic substances. Also, the small particles size of the phosphogypsum
results in a high water content of 20 %–30 %. More importantly, there are also
significant differences between European phosphogypsum and North American
phosphogypsum since they are derived from different phosphate rock.
European phosphogypsum is usually derived from Moroccan 共sedimentary兲 or
Kola 共magmatic兲 phosphate rock whereas Florida rock is the source in North
America. Moroccan rock contains small amounts of radioactive uranium,
which is present in larger quantities in Florida rock. For this reason, great care
must be exercised when interpreting results obtained in Europe.
Many approaches have been investigated to process and utilize
phosphogypsum over the past years. However, most of these have been met
with mixed results at best. It must be concluded therefore that the chance of
phosphogypsum ever being utilized on a large scale is extremely remote. For
this reason they are not discussed further in this publication.
5.3.2 Titanogypsum
This material is available in limited quantities in North America and Europe. It
is obtained by the neutralization of dilute sulphuric acid by calcium carbonate
Calcination Processes 43

under specific conditions such as pH. Characteristically, it has a white color


and a high purity. The only known problem is its fine particle size and needle
like crystal habit that requires blending with other types of gypsum or
agglomeration.
5.3.3 Insoluble Anhydrite
The only material in this category is fluoroanhydrite, which does not require
calcination. Instead, it is neutralized with calcium hydroxide and ground for
use as anhydrite plaster. Activators to promote setting are potassium sulfate,
zinc sulfate together with calcium hydroxide, or Portland cement. These are
usually mixed with the anhydrite at the plant. There are a number of problems
associated with the utilization of this material. It is slow setting despite the use
of activators and it may have a latent acidity, which depends on the manner in
which the occluded fluoric acid is released from the anhydrite crystals.
Fluoroanhydrite also has a different crystal shape from natural anhydrite. It
consists of very small primary crystals that have been combined to secondary
particles with a high surface area and a high reactivity. In contrast, natural
anhydrite consists of very large primary particles, which must be processed
further by grinding.

5.4 Calcination Methods


Little is known about the beginning of ancient calcination methods. The earli-
est descriptions given by the Egyptians refer to the placing of hand-sized pieces
into an open fire. The temperature of these fires could be raised by blow tubes
or bellows. Contamination with ash was minimal since the chunks did not
disintegrate. This method of calcination was changed very little over time and
the Romans appear to have calcined gypsum in a similar manner although
their facilities were larger. It is also not known to what extent they differenti-
ated between limestone and gypsum. There was some improvement, which
probably took place in the 15th century when large pieces of gypsum were
placed on a grate in a furnace to separate it from the ash. It is obvious that for
all these processes, a control over the calcination process was incidental. The
materials produced resembled insoluble anhydrite.
A major departure from the above procedure took place with the advent of
equipment such as the Harz Mountain kettle, probably in the 17th century. This
kettle resembled a shallow pan in which ground gypsum was heated indirectly.
Since this took place under more controlled conditions, it allowed the produc-
tion of stucco instead of anhydrite. Significantly, the calcination process gave
the impression of a mechanism akin to boiling due to the release of steam. This
was an essential feature for more modern processes since release of steam
caused fluidization of the gypsum and maintained a high water vapor pressure
surrounding the individual particles. Modern, larger kettle processes, whether
continuous or not, makes use of this concept.
Calcination is obviously the most critical operation of the gypsum plant. If
the gypsum is under-calcined, the residual gypsum may reduce the setting
times. If present in large enough quantities, it will also lead to lower strength. If
the gypsum is over-calcined to contain soluble anhydrite, the energy consump-
44 Gypsum

Fig. 5.2—Calcination process for grinding before and after calcination: 共a兲 inlet; 共b兲
breaker; 共c兲 transport arrangement; 共d兲 intermediate silo; 共e兲 feeder; 共f兲 magnet drum;
共g兲 hammer mill; 共h兲 silo; 共i兲 transport screw; 共k兲 kettle; 共l兲 hot pit; 共m兲 transport screw; 共n兲
silo; 共o兲 cell feeder; 共p兲 grinding mill; 共q兲 classifiers; 共r兲 large size return; 共s兲 silo; 共t兲
bagger; 共u兲 dust collector; 共v兲 exhaust pipe.

tion is increased and the setting properties of the stucco may be affected.
Chloride levels can affect the calcination of gypsum through a process
known as aridization. This generally results in lower water demand of the
stucco produced and it is more noticeable with batch calcination. For this
reason, it is sometimes referred to as artificial aging. Chlorides can lower the
calcination temperature and thereby result in undesirable phase distributions
of the stucco. Most importantly, chloride salts interfere with the bond between
the core and the paper of gypsum board. Therefore, specifications for chloride
levels are quite stringent. However, efficient washing techniques have been de-
veloped, which can reduce chloride to acceptable levels, usually below
100 ppm.
Although many industrial processes for the calcination of gypsum have
been developed, only a few have been used extensively. They are usually char-
acterized by simplicity of design and by robust equipment combined with low
capital and operating costs. The intention is to manufacture as uniform a cal-
cination product as possible. This calcination product is always based on
stucco, that is, ␤-hemihydrate.
At the present time there are various calcination processes in use in North
America and in Europe. The main processes, however, are based on the Ehrsam
Kettle and to a lesser extent the Rotary kiln. Possible exceptions are systems
that grind and calcine at the same time such as the flash calcining Raymond
impact mill 共imp mill兲 and the Claudius-Peters ring ball mill. Grenzebach BSH
produces similar impact mills and roller mills that are nearly identical to the
Claudius-Peters ring ball mill. The term flash calcining is used to define pro-
cesses that produce stucco rapidly, that is, in terms of seconds or minutes
rather than hours. Similarly, Calcidyne operation involves rapid calcining of
gypsum that is ground prior to feeding.
The conventional kettle and other equipment used for the calcination of
Calcination Processes 45

ground natural gypsum are also suitable for the calcination of FGD gypsum.
Agglomerated material is normally required for units that require coarse natu-
ral gypsum such as rotary kilns, conveyor kilns, impact mills, or roller mills.
Flash calciners without grinding capabilities, such as the Calcidyne and the
Holoflite units, are also suitable for FGD gypsum. Moreover, the Raymond cage
mill, which was converted to a dryer/flash calciner for moist gypsum, has been
used successfully in several plants.
In Europe, continuous and conical kettles with capacities of up to 45 tons
have been use for some time. Some difficulties in conveying and mixing have
been experienced because of the unusual flow properties of the dry FGD gyp-
sum. This can result in material flowing out of gates that retain natural gyp-
sum. Similarly, dry FGD gypsum may consolidate 共bridge兲 and resist mixing
under certain conditions.
Kettle calciners operate most efficiently when the calcining mass is well
fluidized by the steam generated by the calcination process. Often, smooth
fluidization is dependent on a relatively broad particle size distribution. This
can be accomplished by grinding or the addition of ⬍10 ␮m material. The
relationship between size distribution of FGD gypsum and calcination proper-
ties has not been investigated. Consequently, the details of the modifications of
the distribution to improve calcination properties are not known.
A typical arrangement for a calcination process is shown in Fig. 5.2.
5.4.1 Batch Kettle
The batch kettle was initially the process adopted by the gypsum industry for
the production of stucco that was intended for the manufacture of so-called
hardwall plaster. This kettle consists of a steel shell 3 / 8 in. 共0.95 cm兲 thick and
a convex bottom 1 1 / 2 in. 共3.2 cm兲 thick, heated indirectly at the bottom. The
earlier kettles were pierced by four flues about 14 in. in diameter, normally
arranged two above two but sometimes with four in one line. Kettles are
usually about 12⫻ 12 ft2 共3.7⫻ 3.7 m2兲 in size with a capacity of about 20 ton
of gypsum. Other kettles can have capacities as low as 7 – 10 ton and as high as
40– 45 ton. The example discussed below is based on a 20-ton kettle.
Kettles are normally fitted with burners that allow a low and high firing
rate. The combustion chamber should contain most of the flame to reduce
radiation to the kettle bottom. Also, minimum excess air should be used to
accomplish complete and clean combustion. Leakage of air into the system
should be avoided. Baffles are provided in the combustion chamber to direct
hot gases against the bottom. If the equipment is suitable the hot gases should
also pass up the sides, down the back and through the flues of the kettle to the
stack. The heat transfer surface of both the kettle bottom and the cross flue
tubes should be cleaned periodically in order to maintain maximum efficiency.
The fuel used for heating is usually gas but oil and coal may be used as well
共Fig. 5.3兲.
Originally, batch kettles were operated with a calcination cycle of about
2 – 3 h. In this process, the dried gypsum is ground to ⬍2 mm 共0.08 in.兲 and
added to the kettle over a period of approximately 15– 30 min. The stirrer is in
operation at that time and the kettle is heated according to a specific program.
46 Gypsum

Fig. 5.3—Batch kettle with four transverse flues, usually oil or coal fired. 共a兲 Kettle; 共b兲
kettle bottom; 共c兲 furnace; 共d兲 baffles; 共e兲 flues; 共f兲 stack; 共g兲 exhaust lid; 共h兲 exhaust pipe;
共i兲 stirrer motor; 共k, l兲 stirrer; 共m兲 dumping slide; 共n兲 hot pit; 共o兲 transport screw.

The calcination temperature 共drag兲 is about 125° C 共257° F兲 and the dump
temperature is normally about 160° C 共320° F兲. For a hot kettle, filling takes
place immediately following the dumping. The total thermal efficiency 共energy
including electrical energy兲 of batch kettles is low at about 55 %. 共Table 5.1兲
When starting with a cold kettle, heat is supplied for about 3 h using the
low firing setting. This is intended to minimize thermal shock of the refractory
materials within the kettle. While the stirrer is operating, a small sealing charge
of about 2 tons of gypsum is then fed into the kettle and is allowed to come to a
boil with the burner still at a low firing rate. Feeding about 5 tons into the
kettle follows the sealing charge and the burner is set at the high firing rate. The
remainder of the gypsum is then added at a rate to maintain boiling.
At the end of the calcination cycle, the stucco is “dumped” within 30 min
from the kettle into the hot pit, a silo built from concrete or steel. The stucco is
normally cooled to about 125– 140° C 共257– 248° F兲 and post-calcination
grinding often takes place. This is a critical step because the stucco ages on
exposure to water vapor in the hot pit. Also, both the rheological properties
共e.g., consistency or water demand兲 and the setting characteristics can change.
Calcination Processes 47

In extreme situations, liquid water is formed in the hot pit when the water
vapor condenses. This leads to the formation of dihydrate and very short
irregular setting times.
Stirring of the batch kettle contents is provided primarily by the steam
released during the calcination. Additional mechanical stirring is necessary for
the filling and cook-off periods 共time just prior to dumping兲 when the steam
release is slow. The standard mechanical stirring is provided by a central shaft
approximately 1 ⫻ 6 in.2 共2.5⫻ 15.2 cm2兲 and sufficient in length to reach
within less than 6 in. 共15.2 cm兲 of the kettle wall. The blade of the paddle is set
at an angle of 45° so as to plow the material in the kettle upward and towards
the center. Stirring is of particular importance for the bottom arm that has a
different shape and conforms to the bottom of the kettle. The material in this
area must be constantly moved to maintain efficient heat transfer and avoid
overheating. Drag chains are added to the bottom arm for this purpose. There
are numerous arrangements for stirrer and drag chains in use, all intended to
move the calcining material away from the kettle surfaces.
To ensure constant stucco quality, it is essential to maintain all operational
steps as constant as possible. This requires diligent operation if the kettle is
operated manually. However, many of these problems can be overcome by an
automatic operation.
The batch kettle in general has a lower rate of production than the
continuous and submerged combustion kettles discussed subsequently. This is
mainly due to extra time required by the emptying and refilling of the batch
kettle. However, optimization of operation can be achieved by controlling the
following parameters: Stacks, filling systems, kettle top openings, kettle gates,
hot pits, scalping screens, aridization 共lowering of the consistency through
chloride additions兲, and instruments for kettle temperatures, firing control,
control panels, and interlocking systems. In that context, it is most important
to establish a calcination plan with appropriate control functions for such
operating steps as cold and hot start ups, kettle filling control, drag, and cook-
off and dumping. To avoid creating dust, the kettle should not be filled too high.

5.4.2 Continuous Kettles


The conventional batch kettle used to compete with the rotary kiln, particularly
since the latter operated on a continuous basis and provided nearly
instantaneous turn-on and shut-off means. This has changed since the kettle
has been modified to a continuous operation. BPB Canada 共Westroc兲, now
CertainTeed Gypsum, developed the continuous process in the late 1950s and it
is now used almost universally. The key change consisted of the introduction of
mechanisms that allowed simultaneous adding of gypsum and removal of
stucco. This led to an increase in the rate of production and the thermal
efficiency. The increase in the rate of production relates to the batch kettle,
primarily because of time saved through fewer start-up and dumping steps.
About 10– 12 ton/ h can be produced in a 20 ton kettle. Also, continuous kettles
are fitted with up to 18 flues and the gases are passed though the flues more
than once. The partial recirculation of combustion gases tends to increase
48 Gypsum

Fig. 5.4—Continuous stucco production: 共a兲 gypsum silo; 共b兲 drying-grinding unit; 共c兲
cyclone; 共d兲 feed bin; 共e兲 continuous kettle; 共f兲 hot pit; 共g兲 electrostatic precipitator; 共h兲
stucco outlet.

thermal efficiency, which is about 65 %, considerably better than that of the


batch kettle.
The initial start-up for continuous operation is similar to that used for the
batch kettle. This consists of filling the preheated kettle with raw gypsum while
heating and stirring is maintained. When the kettle contents reach the required
control temperature, the addition of the gypsum is balanced with the discharge
of the stucco to the hot pit. During the calcination of gypsum the developing
water vapor fluidizes the material, as in the case of the batch kettle. The stucco
settles to the bottom and it is continuously discharged near the top through a
pipe connected to the inside of the kettle 共Fig. 5.4兲.
The operation of the continuous kettle has a number of advantages over the
batch kettle such as a more constant product, easier operation, less wear, more
efficient utilization of dust collectors, and constant waste heat from burners.
In some plants, the gypsum is ground ahead of the calcination instead of or
in addition to grinding the stucco after calcination. However, these procedures
do not have the same effect. Grinding gypsum before calcination broadens
primarily the particle size distribution of the gypsum in order to change
Calcination Processes 49

TABLE 5.1—Total energy efficiency and rate of production of kettles.


10 ton kettle 20 ton kettle

Operating Energy Energy


Temperature, Efficiency, Production Rate, Efficiency, Production Rate,
°C % ton/h % ton/h
Batch process 125 35–45 3–5 45–65 8–10
Continuous process ⬎140 40–45 4.5–5.5 55–75 9–12
Submerged combustion ⬎140 55–60 6–9 63–83 16–19
Conical kettle 140–170 90 20

Fig. 5.5—Kettle with submerged combustion: 共a兲 stirrer; 共b兲 outer casing; 共c兲 pipe for
stucco discharge; 共d兲 combustion chamber; 共e兲 cross flues; 共f兲 pipe burner for submerged
combustion; 共g兲 recirculation fan.
50 Gypsum

primary the flow properties of the gypsum. Grinding of the stucco produces
new surfaces that can.materially change the rheological and the setting
properties of the stucco.

5.4.3 Submerged Combustion Kettles


This technique has been developed as a means of further increasing the rate of
production of existing kettles. As a result, a number of kettles are now equipped
with submerged combustion providing about 35 % of the energy consumed
共Table 5.1兲. The production rate of both the batch and continuous kettle is
limited by the quantity of heat, which can be transferred to the material to be
calcined mainly through the bottom of the kettle and to a lesser extent through
the flue tubes. Above a given temperature limit, there is the possibility of more
frequent kettle damage.
Therefore, the submerged combustion technique was developed for the
purpose of providing a more efficient method of direct heat transfer. The
obvious advantage of the process is that extra heat can be supplied directly to
the material being calcined without raising the bottom temperature of the
kettle.
The material in the kettle is fluidized primarily by the steam generated
during calcination. The heat transfer in a fluidized bed is very high and the heat
contained in the combustion gases is passed directly to the kettle content.
Therefore, the submerged combustion system increases the fuel efficiency of
the continuous kettle. The efficiency is increased to approximately 75 %
excluding the normal bottom burner. The gases leave the bed about 5 – 10° C
共41– 50° F兲 higher than the kettle control temperature, which contributes to the
high thermal efficiency of the system. When using natural gas, the burner is
adjusted to operate with only 10 % excess air, which also adds to the fuel
efficiency.
Typically, when an existing kettle is fitted with a submerged combustion
unit, there is also a saving in electrical energy in addition to the increase in rate
of production. When using submerged combustion, the amount of dust passing
the dust collector is obviously increased. To reduce the dust loading to the
collector, a level control device is used for the kettle as a standard feature 共Fig.
5.5兲. When properly operated, the quality of the stucco produced changes little.
However, there is some reduction of setting times, which is an advantage if one
considers the manufacture of gypsum board.

5.4.4 Conical Kettle


The conical kettle is the logical extension of the experience obtained with the
continuous kettle having submerged combustion. The kettle consists of a
cylinder with a conical bottom and a cylindrical upper portion. All the calcining
heat via the combustion gases is passed directly into the calcining material.
This is done from the top of the kettle through a tube reaching almost to the
bottom, which is protected from the heat of the flame by a ceramic or concrete
cone. Gypsum is added continuously to the top of the kettle and stucco is
removed through an overflow arrangement located at the conical part of the
kettle. Significantly, the conical kettle does not have a stirring mechanism.
Calcination Processes 51

Fluidization and stirring of the calcining material is achieved by a combination


of steam being generated during calcination and the flue gases being supplied
by submersed combustion. This mixing effect is assisted by the conical bottom
of the kettle 共Fig. 5.6兲.
The greatest savings in energy consumption is claimed for this process
because of the direct contact of all of the flue gases with the calcining material
as well as the absence of an electrical mechanism for stirring. In this manner, a
20 ton kettle produces about 20 ton/ h of stucco with an energy efficiency of 90
%. However, care must be exercised for this process since the material is
exposed directly to the heat of the burner flame operating at a relatively high
temperature. For all kettles, it is characteristic that at the end of the normal
calcination procedure both soluble anhydrite and dihydrate are present in
small amounts. This can be exacerbated in the conical kettle. Moreover,
insoluble anhydrite can be formed in substantial quantities when this kettle is
being operated incorrectly.
Complete and accurate tests are essential for all kettle operations to
develop and maintain good calcination practices. A testing program should be
developed for each particular plant to show the properties of materials at all
stages of the process. However, the complete control of quality requires more
than just the testing itself. It requires intelligent interpretation of the test
results to determine where and when changes should be made. Therefore, other
observations of the performance of the entire system should be taken into
consideration 共Table 5.2兲.
5.4.5 Rotary Kilns
A typical modern rotary calciner is basically a ceramic lined steel cylinder, 8 ft
共2.4 m兲 in diameter and 130– 155 ft 共39.6– 47.2 m兲 long that is installed at a
slope of 1 / 2 in. per foot 共3.9 cm per meter兲. The speed of rotation is about
1 – 2 rpm. In North America, direct heat is normally supplied in a counterflow
manner, that is, with the heat entering the shell at the stucco discharge end.
The feed for the rotary kiln is relatively coarse with a size range of 0 – 1 in.
共0 – 25 mm兲, which often does not have to be dried prior to calcination. This
feed is heated by direct contact with the flue gas as well as by radiation from
the shell. The internal design of the shell must provide uniform distribution of
feed and heat supply. In addition it must provide sufficient air to allow proper
combustion and to remove the steam released during calcination.
Earlier versions of the rotary kiln calciners were shorter and did not
contain distribution diaphragms. They were used with co-current flow by
introducing the feed and the heat at the front end. Difficulties such as
maintaining a uniform flue temperature led to changes in design that are
commonly used in North America. In Europe, co-current rotary kilns, as shown
in Fig. 5.7 are still in use.
Rotary kilns are more energy efficient than average size batch kettle or
simple continuous kettles with a capacity of 20 ton/ h. Continuous kettles with
a capacity of 40– 45 ton/ h are almost equal to rotary kilns and only conical
kettles are more efficient. Primarily, this is due to the higher production rate of
kilns producing up to 25 ton/ h of stucco and the corresponding fuel savings.
52 Gypsum

Fig. 5.6—Conical kettle.

For these reasons and because the kiln has almost instantaneous start-up and
shut-down capabilities, rotary kilns were at one time the calciner of choice for
some companies in North America.
Grenzebach BSH manufactures a co-current rotary tubular kiln, which is

TABLE 5.2—Analysis of stuccos produced.


Soluble Insoluble Water Setting,
Dihydrate, Anhydrite, Anhydrite, Demand, Time
% % % mL/ 100 g min.
Batch kattle 3 11 2 64 25
Continuous kettle 4 6 2 65 18
Submerged combustion 4 6 2–4 65 15
Conical kettle 4 6 3–4 65 15
Calcination Processes 53

Fig. 5.7—Co-current rotary kiln.

fitted with several flues along the length of the kiln to increase the heat
exchange. This type of kiln requires a feed of about 2 mm 共0.08 in.兲 instead of
the usual 25 mm 共1 in.兲.
If properly operated, the rotary kiln can produce stucco similar in
uniformity and quality to kettle stucco, both for the manufacture of hardwall
plaster and gypsum board. This can be achieved only by feeding the proper size
gypsum at a uniform rate through the rotary kiln and by heating the gypsum
uniformly. Various mechanical methods are used for this purpose. For example,
a diaphragm is used to spread out the feed that may be damp and to provide
more surfaces for heat transfer.
However, since the feed to the rotary kiln consists of particles ranging in
size from fine dust to about 1 in., there is a tendency to over-calcine the fine
dust to a water-of-crystallization content as low as 1 %. Conversely, the centers
of the very large particles tend to be under-calcined to a water-of-crystallization
content as high as 10 %. A feed with a fines content of less than 10 % through a
100-mesh screen is satisfactory because it tends to roll over itself while passing
through the shell. This brings the coarse particles to the top and protects the
finer particles in the center.
The dust from the collector is the least desirable part of the product. It
forms a significant part 共5 %–10 %兲 of the stucco produced. Since the dust is
collected from the entire length of the calciner, it contains a mixture of gypsum,
hemihydrate, and soluble anhydrite. The material is further calcined by the hot
gases that carry it to the collector. Consequently this material is usually soluble
anhydrite and insoluble anhydrite. Insoluble anhydrite remains unchanged and
constitutes an impurity, which may be significant. Also, the dust can have a
high pH because of the possible decomposition of the insoluble anhydrite to
CaO and SO3.
FGD gypsum must be agglomerated prior to calcination using this type of
calciner. Rotary kilns that use a powdered gypsum feed are not common, except
the tubular rotary kiln produced by Grenzebach-BSH. A possible advantage of
such a technique is that drying and calcining may be carried out in one step.
54 Gypsum

Fig. 5.8—Imp mill calcination.

5.4.6 Impact Mill Calciner


Some flash calciners 共e.g., impact mill and ring ball兲 grind and calcine the
gypsum simultaneously. The feed for these operations is generally in the order
of 3 / 4 in. 共1.9 cm兲. Agglomerated FGD gypsum can be used for these calciners,
which has the added advantage of modifying shape and size distribution.
Alternatively, the hammers can be removed from the imp mills to process the
FGD gypsum in a dry or wet state. For either approach, calcination parameters
and the effect of size distribution require further examination.
A well known flash calcination system in North America is based on the
impact hammer mill 共imp mill兲. It grinds and calcines gypsum simultaneously
using the same Raymond impact hammer mills that are used to pulverize
Calcination Processes 55

gypsum. The large version of this mill produces about 12 ton/ h with a fineness
of 96 % passing a 100-mesh 共150 ␮m兲 screen or ⬍0.2 mm. The gypsum is
generally pre-ground to pass 3 / 4 in. 共1.8 cm兲 screen. A smaller version of this
mill is in use producing approximately 2 1 / 2 ton/ h or less.
Hot gases, generated in a burner outside the mill, are blown into the
grinding chamber of the mill while the gypsum is being pulverized. The ground
and calcined material is carried out of the mill by the hot gases and collected in
cyclones. As in conventional calcining, the hot stucco is transferred to a hot pit
for cooling. Attempts have been made to make this a truly start-up and shut off
process by avoiding the hot pit. This has not been successful since the hot pit
appears to be needed to produce uniform stucco by calcining residual gypsum
and converting soluble anhydrite to hemihydrate. Very fine material remaining
in the gas stream is removed by bag houses and combined with the calcined
gypsum. The hot gases are recycled to the burner, which is one reason for the
relatively high-energy efficiency of this process 共Fig. 5.8兲.

5.4.7 Ring Ball and Roller Mills


These flash calcining systems originated in Europe. The Claudius-Peters ring
ball mills are similar to roller mills except that balls are used instead of wheels
as in the Grenzebach-BSH calciner. Both are often used for simultaneous
grinding and drying of coal. Therefore, the move to a combined grinding and
calcining of gypsum was a logical step. The operation of the ring ball grinder/
calciner is very similar to that of the imp mill calciner, except that gypsum up to
1 1 / 2 inch 共3.8 cm兲 can be used as feed material. The stucco produced has a
particle size of approximately ⬍0.2 mm 共100-mesh or 150 ␮m兲 共Fig. 5.9兲. Exact
thermal and electrical energy requirements are difficult to establish. Thermal
energy consumptions of similar devices were found to be between 940 and
1312 kJ/ kg 共890,000⫺1,234,000 Btu/ ton兲. The corresponding electrical energy
requirements are generally between 34 and 53 kWh/ ton.

5.4.8 Calcidyne Unit


This flash calciner accepts pulverized feed intended for a single pass. It
resembles a vertical rotary kiln, directly heated, with baffles to control the
movement of material through it. The gypsum and flue gases enter the unit at
the bottom and the stucco is collected at the top. The flue gases and the water
vapor from calcination are recycled into the bottom of the calciner. The
Calcidyne unit is suited for FGD gypsum because it accepts material with a
narrow feed size distribution 共Fig. 5.10兲.
There are various other calcining units in use in North America and
elsewhere, such as the Holoflite. These are not discussed further here because
of their similarity to other units or because they are not significant.

5.4.9 Anhydrous and Multiphase Plaster


Anhydrous plasters are produced with dry calcining processes between 300 and
900° C 共572 and 1652° F兲. Usually three reaction stages are being considered,
i.e., ⬍500° C so-called slowly soluble anhydrite, 500– 700° C 共932– 1292° F兲
insoluble anhydrite and ⬎700° C floor screed plaster. Specific ratios between
56 Gypsum

Fig. 5.9—Claudius-Peters ring ball calciner with classifier.

the reaction stages are required, which are determined by both the raw
material and the calcination process. Finally, these plasters are mixed with
conventional stucco to obtain multiphase plasters.
A modern calcining process for anhydrite plasters is the conveyor kiln.
Capacities of 1200 tons per day are usual and the thermal efficiency of the
conveyor kiln is believed to be greater than 70 %. Before being fed into the
conveyor kiln, the gypsum is crushed to pass a 60 mm 共2.4 in.兲 screen and split
into several sieve fractions, i.e., 7–25, 25–40, and 40– 60 mm 共0.2–1, 1–1.6, and
1.6– 2.4 in.兲. The fractions are placed in layers onto a moving grate, the
smallest size on the bottom. The grate moves at a speed of 20– 35 m / h
underneath a calcining hood from which the hot gases are drawn through the
gypsum by fans. There is little dust being created because the gypsum is not
moved during calcination and therefore no dust collection is required. About
half of the gases at approximately 270° C 共518° F兲 are recirculated to the
combustion chamber and mixed with air at about 230° C 共446° F兲, which is
then drawn through the plaster bed for cooling 共Fig. 5.11兲.
For the production of multiphase plaster, intended for spray application,
the anhydrite is mixed in approximately equal portions with conventional
stucco. This stucco is produced in a rotary kiln using the 0 – 7 mm fraction,
Calcination Processes 57

Fig. 5.10—Calcidyne calcining system.

which is subsequently ground before being mixed with the anhydrous plaster.
The multiphase plasters produced differ in their setting properties from each
other as well as from stucco.
Multiphase plaster intended for spray application has been produced since
1965 by incorporating chemical additives. These spray applied plasters differ
from similar North American plasters because they are mixed with water at the
spray nozzle rather than being mixed with water and pumped prior to spraying.
Multiphase construction plasters without additives 共Saarland Construction

Fig. 5.11—Conveyor kiln for anhydrite production: 共a兲 conveyor grate; 共b兲 feed hoppers;
共c兲 calcining hood; 共d兲 gypsum; 共e兲 combustion chamber; 共f兲 discharges; 共g兲 recirculation
and cooling fans.
58 Gypsum

Plaster, Southern German Double Calcined Plaster兲, which are prepared in


these types of plants, have declined in use recently.
Other calcining plants produce anhydrite in a counter current rotary kiln
共Vernon兲. For example, Surcuit is produced in France as Paris Building Plaster
by mixing anhydrite with stucco. In this type of kiln the gypsum reaches a
maximum temperature of about 500° C 共932° F兲 just before discharge. These
kilns, with capacities of 15– 30 ton/ h, are sometimes linked together to
produce first anhydrite followed by the production of stucco. This improves
their thermal efficiencies since it provides for cooling of the anhydrite. The
problem of cooling has also been solved by using the planetary cooling system
of the cement industry.
Anhydrite can also be produced by grinding natural anhydrite rock in tube
mills or impact mills to a particle size of below 0.2 mm. Activators to promote
setting are added together with the gauging water. These activators are
mixtures of alkali metal or heavy-metal salts and calcium hydroxide. They are
being used in quantities of up to 2 % by weight based on the anhydrite. Acid
activators such as potassium hydrogen sulfate or iron 共II兲 sulfate can also be
used.
5.4.10 ␣-Hemihydrate Plasters
Calcination by steam autoclaving originates with the 1933 patent by Randel
and Dailey 关5-2兴. In this patent the term alpha-gypsum was coined and that
term was adopted in the scientific literature as ␣-hemihydrate. This led to the
term ␤-hemihydrate produced either in the laboratory or as stucco in a kettle.
In essence, the patent describes the process of calcining gypsum in an
autoclave for approximately 6 h at a temperature of 235° F 共113° C兲 and a
pressure of 15– 17 psi. The resulting plaster was capable of producing a cast of
gypsum with unusually high strength due to the low water demand of
42– 45 mL/ 100 g. This process became the basis for ␣-plaster production in
North America.
For this process, gypsum rock, 1 / 2–2 in. 共1.27– 5.08 cm兲 in size, was
preferred because of ease of handling and drying. Because of the size of the
material used, it is sometimes called the lump process. It is also called the dry
process to distinguish it from the slurry autoclaving or solution processes
developed at a later time. The patent specifically states that ground gypsum or
synthetic gypsum may be briquetted and used as well. This concept has been
developed further in Europe, for example, by autoclaving pressed bricks of
FGD gypsum using sand/lime brick manufacturing facilities.
The first practical batch process for slurry autoclaving was developed by
Schoch and Cunningham in 1940. They found that autoclave calcination in
concentrated magnesium chloride solutions produced short rod-like crystals
having a water demand of 45 mL/ 100 g. The water demand could be reduced
by ball milling to 35 mL/ 100 g. A pilot plant based on this process established
the cost for commercial production as US$8.82 per short ton.
Haddon and Cafferata describe the use of crystal habit modifiers such as
succinic acid and a relatively low temperature of about 120° C. Commercial
processes have been developed on the basis of this invention which typically
Calcination Processes 59

Fig. 5.12—Autoclave for ␣-plaster production.

produces ␣-plaster with a water demand of less than 35 mL/ 100 g.


A continuous autoclave process developed by Giulini is aimed at synthetic
gypsum from the phosphoric acid manufacture. This process uses salt solutions
and crystal habit modifiers as well as a relatively low temperature and a low
pH. The process was not entirely successful mainly because of the isomorphic
phosphate impurities, which interfere with subsequent gypsum crystallization.
Another continuous autoclave process has been developed by Knauf using
FGD gypsum 关5-3兴. Calcination takes place in the presence of crystal habit
modifiers at approximately 125° C and at 2 – 3 bar pressure. The ␣-plaster
produced has a low water demand and is suitable for floor screed applications.
Production costs for this type of process are usually given as about $60 per
short ton, mostly because of high-energy costs 共Fig. 5.12兲.
Finally, Eastroc in Pennsylvania, a subsidiary of BPB North America, now
CertainTeed Gypsum, had developed an elegant and interesting process, which
uses an autoclave to convert the calcium sulfite from the scrubber directly into
␣-hemihydrate. To a large extent this process utilizes the exothermic condition
of this reaction. At the present time this plant is not in production 关5-4兴.
Considering the above mechanisms it becomes obvious that ␣-hemihydrate
consists of fairly large, single crystals that have a low aspect ratio. The essential
key is a low water demand and, consequently, a higher strength. In most cases
the ␣-hemihydrate plaster has different rheological properties from regular
plaster based on stucco. Calcining “lumps” or briquettes in an autoclave can
produce this material. Alternatively, it may be obtained in slurry-autoclaves or
with slurries at atmospheric pressures using inorganic or organic crystal habit
modifiers. If these crystals are needle-like they produce a high water demand.
60 Gypsum

Under certain conditions it may be possible to achieve a low water demand by


subsequent grinding all or part of the material 关5-5兴.

References
关5-1兴 Jurkowitsch, H. and Hüller, R., “Handling and Processing FGD Gypsum for Appli-
cation in the Building Industry in Europe,” Fourth International Conference on
FGD and other Synthetic Gypsum, May 17–18, 1995, Ortech International, Toronto.
关5-2兴 Randel, W. S. and Dailey, M. C., “High Strength Calcined Gypsum and Process of
Manufacturing Same,” U.S. Patent 1,901,051 共March 14, 1933兲.
关5-3兴 Wirsching, F., “Production and Use of Industrial Calcium Sulfates and the Proper-
ties of Building Material Produced from Them.” ZKG Int., Vol. 44, 1991, pp. 248–
262.
关5-4兴 College, J., “The Omega Process: Direct Conversion of Calcium Sulphite to Alpha
Hemihydrate,” Sixth International Conference on Natural and Synthetic Gypsum,
May 16–19, 2000, Ortech International, Toronto.
关5-5兴 Kuntze, R. A., “Alpha-Plaster from FGD Gypsum and Potential Markets,” Fourth
International Conference on FGD and other Synthetic Gypsum, May 17–18, 1995,
Ortech International, Toronto.
MNL67-EB/Oct. 2009

6
Properties of Stuccos and Plasters
6.1 Stucco Requirements
STUCCOS PRODUCED BY ONE OF THE CALCINATION PROCESSES ARE
generally formulated by adding a variety of materials at the mill to control or
adjust their properties. Formulation usually means that at least retarders are
added to lengthen the set time of the plaster. Often, stabilizers, such as a com-
bination of aluminum sulfate 共papermakers’ alum兲 and hydrated lime, are
added to make the setting of the plaster less sensitive to freshly set gypsum or
other accelerators inadvertently added at the job site. Other additives are used
to adjust various functions of the stucco, such as water retention. If lightweight
aggregates, such as expanded perlite, are involved, they are usually added at the
mill as well. After the addition of these materials, stucco is called 共dry兲 plaster.
The dry plasters are then normally mixed with sand and subsequently with
water at the job site to be used as 共wet兲 plasters for specific applications. At this
stage the wet plaster resembles hydrated lime or cement mortar, which have a
similar historical development to gypsum plaster. In North America, the prime
example is hardwall plaster, which is stucco that is retarded to about 3 – 5 h. It
is also stabilized so that freshly set gypsum has a minimal effect. This material
is applied in two or three coats to a plaster base, such as gypsum lath that is
fastened to studs. Finally, a neat 共not sanded兲 finishing coat consisting of a
mixture of stucco and hydrated lime is applied over the last coat. As mentioned
before, this process has been largely replaced by “drywall” applications.
The corresponding material in Europe consists either exclusively of regular
stucco or of finishing plaster, which is a combination of insoluble anhydrite and
stucco. Finishing plaster differs from regular stucco in having a lower water
demand and a more gradual setting behavior. Specifically, this means that it has
an earlier initial set and a later final set. Dry plasters are prepared from these
materials by the addition of retarders, fillers such as limestone, or aggregate
such as expanded perlite. To prepare mortars, sand is added at the job site.
Finishing plaster is primarily used for spray application by machine. All other
plasters are usually applied by hand.
Other requirements for stucco are concerned with the minimum hemihy-
drate content as well as proper shipping and storage conditions. Additional
requirements for plaster are concerned with number of coats, the effect of frost,
and the absorption capacity of the plaster base.

61
Copyright © 2009 by ASTM International www.astm.org
62 Gypsum

For board manufacture, it is common practice to require that the stucco


meets certain rudimentary particle size specifications. For stucco from natural
sources these are generally expressed, for example, as 90 % passing a 100-mesh
共150 ␮m兲 sieve. This takes into account a specific knowledge of the grinding
processes that are used for both the gypsum prior to calcination and for the
stucco by the so-called post-calcination grinding. In Europe, the specification
calls for regular stucco to retain 10 % on a 0.2 mm sieve 共approximately 100-
mesh兲, and 40 % for finishing plaster.
The fineness of the gypsum can be significant because it affects the degree
of calcination. Coarse gypsum may possess a high amount of insufficiently
calcined material particularly in the center of large particles. Conversely, the
exposed surfaces of some particles may be over-calcined, that is, it may contain
soluble or insoluble anhydrite. Fine gypsum may be successfully calcined but
the water demand of the board slurry may be increased. This may adversely
affect the operation by requiring slower machine speed and more fuel to dry
the board.
For single charge kettles and Holoflite units, the consistency of stucco is
sometimes controlled by the addition of calcium chloride, the so-called aridiza-
tion process. In general this lowers the water demand of the stucco by a mecha-
nism akin to aging, i.e., by the prevention of disintegration. Caution must be
exercised since chloride salts have a detrimental effect on gypsum crystal
growth when used above 2 lb per ton of stucco produced.
Further grinding is sometimes required for FGD gypsum if these specifica-
tions are to be met. Grinding is necessary for materials that have a rod-like
crystal shape and a bulk density of about 0.5 ton/ m3 共31 lb/ ft3兲. Because of its
small particle size of 20– 60 ␮m 共maximum 200 ␮m兲 FGD gypsum is preferen-
tially calcined only in kettles unless it is agglomerated. This type of kettle
stucco may be difficult to use as plaster since it has a tendency to demonstrate
peculiar consistency behaviors sometimes mistaken to be thixotropic. It may
also lack plasticity, smoothness, and a volume yield that is too variable. Grind-
ing is not required for materials with a cubic crystal shape and corresponding
bulk density of about 1.2 ton/ m3 共75 lb/ ft3兲.
Aging is a factor that lowers the consistency whereby the rate and degree of
this effect depends on the amount of water vapor absorbed by the stucco 关6-1兴.
This in turn depends on the water vapor pressure 共relative humidity and tem-
perature兲 to which the stucco is exposed. It is significant that the consistency
can be lowered rapidly within a day when the amount of water vapor absorbed
is high, a condition which can occur in the hot pit. It should also be noted that
the stucco continues to age once water vapor is absorbed. For this reason,
plaster placed in bags continues to age if it was previously exposed to water
vapor. The aging process can be terminated by drying but it is not possible to
reverse this process.
For board manufacture, this property is important since under certain con-
dition the stucco may remain in the storage bin for longer than intended. This
can lead to a variety of situations and it is an example of how non-uniformity
occurs. If the stucco remains hot, it is possible that soluble and even insoluble
Properties of Stuccos and Plasters 63

anhydrite is formed with the dihydrate being converted to hemihydrate. Under


certain circumstances, prolonged storage is used to “finish off” the calcination
resulting in a much slower set on the board machine.
Under normal conditions the stucco will cool and it will do so unevenly.
This creates its own set of problems. Usually these bins are not properly venti-
lated so that the moisture 共either present or generated兲 will condense at the
coolest part of the bin to form dihydrate. However, the main effect of cooling in
the bins is that the water demand of the stucco will be lowered drastically due
to rapid aging. This often produces a slow set and generates peelers or splitters.

6.2 Wet Plaster Properties


The application of these materials as wet plaster involves certain plaster prop-
erties, mainly normal consistency, setting time, and volume change on setting.
These properties influence each other to some extent.
6.2.1 Normal Consistency
This property of stucco is often referred to as water demand. It is usually
expressed as milliliter of water per 100 g of stucco and is indicative of the water
to stucco ratio, which produces a wet plaster with “normal” flow properties.
This characteristic is intended to reflect the workability of the wet plaster and it
is perhaps the most important and least understood of all plaster properties. It
is usually measured by one of several methods such as the depth of penetration
of a conical plunger into the plaster 共ASTM C472兲1 关6-2兴, by placing plaster in a
cylindrical mold and measuring the decrease in height when the cylinder is
lifted 共British Method兲, or by measuring the spread of plaster under its own
weight after pouring 共US Industrial Method兲. An interesting method is used in
the German DIN 1168 procedures where stucco is added slowly to a fixed
amount of water. The end point is reached when the water no longer rises to
wet the plaster by adsorption. In this case, the normal consistency of plaster is
expressed as a gram of stucco per 100 mL of water. For this test, it has not been
established to what extent disintegration takes place.
The workability of plaster is usually considered without reference to its
rheological or force/flow relationship. For example, it is seldom appreciated
that the normal consistency of plaster represents only one aspect of the non-
Newtonian force/flow curve, that is, the yield point. The examples given in Fig.
6.1 illustrate the behavior of plaster, which is characterized by the non-
Newtonian curve. This curve indicates that a certain force is required before a
flow is achieved and that this force corresponds to the yield point of the plaster.
Once flow is attained, the curve beyond the yield point may be straight or it
may have some other configuration that are indicative of other flow
characteristic of the plaster. In contrast, the Newtonian curve starts at the
origin and continues in a straight line.
Another way of demonstrating this difference is by placing a cylindrical
weight on the surface of a Newtonian substance, petroleum grease for example,
and of a non-Newtonian substance represented by plaster. In the first case, the
1
Standard Test Method for Physical Testing of Gypsum, Gypsum Plasters and Gypsum Concrete.
64 Gypsum

Fig. 6.1—Schematic force/flow diagram.

weight will penetrate immediately into the grease and will continue to do so at
a slow but constant rate. For most plasters, the cylinder will rapidly penetrate
the plaster to a certain depth and stop at that level provided the water. Solid
ratio is appropriate.
These conditions are in reasonable agreement with the characteristics of
hardwall plaster as experienced in practice. That is, on the trowel, the plaster is
in a non-flow position because no force is applied and therefore it remains
below the yield point. However, as soon as a certain manual force is applied
through the trowel, the plaster begins to flow because the yield point is
exceeded. When the plaster is applied to the wall and the force exerted by the
trowel is discontinued, the plaster returns to its non-flow position. For this
reason, the plaster does not “slough off” the wall and this is obviously an
important characteristic. It also may give the erroneous impression that plaster
is thixotropic. Similar conditions exist during spray application.
It should be realized that the consistency 共water demand兲 and other
rheological properties of plaster depends primarily on the particle size
distribution of the stucco. In the past, this was a difficult subject because there
was apparently no simple correlation between these two parameters. For
example, grinding of relatively uniform stucco can lower the consistency
instead of raising it. This is due to the production of small particles that fill the
interstitial spaces between the larger particles. Therefore, attempts to use
density instead of particle size distribution give only a first approximation of
the situation. Other factors, such as the surface conditions of the stucco
particles, and substances, such as surface-active agents, do play a role,
sometimes to a significant extent. In addition, consistency is affected by aging,
that is, of decreasing with time after absorption of water vapor.
The understanding of these factors have changed with the work of Lane
关6-3兴 who showed that exposure to liquid water in conjunction with mechanical
action produced the expected correlations between particle size distribution
Properties of Stuccos and Plasters 65

Fig. 6.2—Particle size of stucco before and after disintegration.

and consistency. This means that particle size distributions carried out in
liquids other than water produce different particle size distributions that are of
limited usefulness in this context 共Fig. 6.2兲.
Specifically, the work of Lane shows that mechanical mixing of the stucco
with water for 60 s produces complete disintegration, which changes its
median particle size distribution from 24 to 1.4 ␮m and the Blaine surface
area from 3800 to 27,000 cm2 / g. This has the effect of changing the pouring
consistency from 90 to 210 mL per 100 g, a very high value. Further work by
Wirsching also showed that the largest size fraction of the stucco disintegrates
preferentially whereas the smaller size fractions remain largely unchanged.
Of course, this indicates only the potential of disintegration since complete
66 Gypsum

Fig. 6.3—Methods for the determination of the rate of setting.

disintegration is seldom if ever reached under practical conditions. However,


mixing of plaster always produces some disintegration. For any practical
application, the degree to which this occurs should be established.
6.2.2 Setting Characteristics
Plaster dissolves and precipitates to form a solid mass of interlocking gypsum
crystals, provided its consistency is within a normal range. This process is
known as setting and it must be kept in mind that the actual process is affected
to some extent by the type of rock and the calcining procedure used. It is,
however, possible to follow its progress by measuring the degree of gypsum
formation by a variety of methods 关Fig. 6.3兴. The most practical way is to
measure the heat that is released during hydration. These methods have the
advantage that they can produce a continuous temperature-time curve from
start to finish. They have the disadvantage that cooling affects the temperature-
time curve somewhat. It is also possible to measure the percentage of gypsum
formation by x-ray diffraction, differential thermal analyses 共DTA兲, or
differential scanning calorimetry 共DSC兲, or by weight loss determinations,
which necessitate the removal of samples at intervals.
The rate of hydration can also be followed by mechanically measuring the
degree of hardening or solidification. This is a relatively slow process and
requires that the stucco and water is in the proper proportions, that is, be at
normal consistency. For this purpose, the Vicat needle 共ASTM C472兲 is
generally employed. In some plants the Gillmore needle 共ASTM C266兲2 is the
preferred procedure. A 1 / 4 lb 共113 g兲 needle indicates the initial set on the
board machine at which the board can no longer be disturbed. This means that
all mechanical action such as forming the edges with shoes must take place
before the initial set is reached. A 1 lb 共454 g兲 is used to determine the final set,
which indicates that the board is sufficiently solidified to be cut by the knife.
For gypsum board manufacture, the temperature of the stucco slurry is
significant. In general, the fastest final temperature rise is achieved at
2
Standard Test Method for Time of Setting of Hydraulic-Cement Paste by Gillmore Needles.
Properties of Stuccos and Plasters 67

temperatures between 100 and 120° F 共38 and 49° C兲. Higher or lower
temperatures result in appreciably lower setting times. Maintaining the stucco
slurry temperature at the appropriate level is not an easy process. It must take
into account the cooling rate of the stucco, the water temperature, etc. In some
plants the water can be heated particularly during the winter months. It is
doubtful however whether water should be used to control the stucco slurry
temperature other than by a small degree.
The characteristics of hydration or setting have been studied extensively
over the years. One of the goals of this effort was to arrive at a formula for a
sigmoid curve that would express the mechanism of the conversion of
hemihydrate to dihydrate. An example of this attempt is the work of Schiller
who assumes the existence of a classical sigmoid curve because the maximum
rate of hydration occurs at 50 % 关6-4兴. On the other hand, Ridge assumes that
the maximum hydration rate occurs at 60 % 关6-5兴.
In the past, much of the stucco produced in kettles or kilns was used for the
manufacture of so-called hardwall plasters, that is, plasters for application over
lath or another plaster base. For this purpose, the stucco was cooled to room
temperature and ground in a tube mill before adding the required additives and
bagging the material. Tube milling improves the plasticity of the stucco when
mixed with water. It should also be noted that grinding before or after
calcination can lead to a reduction in water demand, for example, for a plaster
with relatively coarse and uniform particles. Sand is normally added to the
plaster at the job site. Instead of sand, lightweight aggregate such as perlite can
be added at the mill.
As shown in Fig. 6.4, the setting time of hardwall plaster as determined by
the temperature method is increased almost exponentially by the addition of
retarder. In this condition, its setting time is very sensitive to the addition of
accelerators such as freshly set gypsum occurring at the job site. Therefore,
hydrated lime and aluminum sulfate are introduced that provide a relatively
stable setting time in the order of 3 – 4 h. Gypsum itself is unsuitable for this
purpose because its acceleration effect decreases with time and particularly
with exposure to water vapor.
6.2.3 Volume Changes
These effects refer to expansion, shrinkage, and creep during the hardening
process. All cementitious material experiences some volume changes that
depend on the temperature and the condition of the material itself. For plasters
an average thermal coefficient of 20⫻ 10−3 K−1 can be expected. This compares
with a coefficient of about 10⫻ 10−3 K−1 for Portland cement concrete.
Practical experience has shown that no problem exists for gypsum despite the
larger value. However, a sudden temperature shock, such as exposure to
freezing conditions, must be avoided.
Expansion is generally considered to be due to the growth of gypsum
crystals in the plaster matrix but it does not include volume changes during the
very early stages of the setting process. Expansions of 1 mm/ m are possible
under unrestrained conditions depending on the basic composition of the
plasters. Shrinkage is expected to occur because of the water loss through
68 Gypsum

Fig. 6.4—Set time stabilization by combined retarder, hydrated lime, and alum.

evaporation after the plaster has set. This value is 0.3 mm/ m according to
practical experience. It appears that no damage is usually caused by the
processes of expansion and shrinkage while the plaster is still in the plastic
state. These processes are essentially completed when the plaster reaches its
final hardened condition. Occasionally severe problems of expansion have been
experienced for example when the setting takes place in two stages whereby the
second setting occurs in a partially hardened plaster, such as floor screed.
This high degree of tolerance to expansion and shrinkage is obviously due
to the larger pore volume and the ability of the plaster to creep. Creep occurs
under conditions when the plaster is exposed to slow lateral loads particularly
in the presence of adsorbed moisture. In fact, it is the main reason why few
cracks appear in the plaster in practice when the load is applied slowly.

6.3 Properties of Hardened Plaster „Gypsum…


A key consideration is the type of internal bond formed when plaster solidifies
during the process of dihydrate formation. These bonds represent three mecha-
nisms: 共1兲 Formation of a true bond between crystals when ions are shared,
such as twinning; 共2兲 crystals growing past each other into available spaces and
forming a mechanically interlocking network; and 共3兲 hydrogen bonding be-
Properties of Stuccos and Plasters 69

Fig. 6.5—Effect of water on compressive strength.

tween crystals where they touch. Mechanical interlocking is similar to the keg-
of-nails effect where the nails “agglomerate” by friction. However, the gypsum
crystals grow into empty spaces and therefore cannot be disentangled as in the
case of nails. It has also been speculated that the loss of hydrogen bonds are the
reason for the relatively large loss of strength of gypsum plaster, up to 66 %,
when exposed to liquid water.
It should also be taken into consideration that sand or other aggregate is
usually added to the plasters to be applied in the form of mortars. This method
employed first by the Romans using several plaster coats for the base of paint-
ing and frescoes. The Roman approach of using decreasing particle size and
decreasing amount of sand for successive coats towards the surface is not fol-
lowed in North American practice. Instead the same type and amount of sand is
used for each successive layer. In any case, this means that the physical prop-
erties of the sanded plaster can be significantly different from those of neat
plaster. This is the case particularly when the amount of sand added exceeds
specifications or is too fine, which is not unusual.
For the manufacture of gypsum board it is obviously an essential feature
that the gypsum crystals bond to the paper fibers of the liner. A variety of
hypotheses are offered to explain this mechanism. For example, they attempt to
explain the role of the starch, which was originally not necessary when the
board was dried slowly.

6.3.1 Drying Conditions


It is often not realized that the physical properties of hardened plaster are
mainly dependent upon the amount of water that evaporates after setting. In
that condition, the set plaster 共gypsum兲 behaves quite different to non-porous
or solid gypsum found in nature. This effect is relatively large since the amount
of water evaporating after setting is also relatively large. For example, wet
gypsum has less than 60 % of the strength of dry gypsum 共Fig. 6.5兲. In
comparison, Portland cement loses only about 10 % of its strength on wetting.
The plaster as applied to walls and ceilings normally requires several days to
70 Gypsum

Fig. 6.6—Relationship between apparent density and consistency.

dry. In practical terms, this means that the physical properties of the plaster,
such as strength and hardness, improve considerably during this process.
Prolonged exposure of set plaster to water, such as delayed drying, should
be avoided since it may lead to permanent changes of the structure of the
plaster due to recrystallization of the gypsum. This condition, called brownout
in the field, is extremely difficult to reproduce in the laboratory. This picture is
being complicated by the fact that subsequent drying will often restore the
original physical properties in many instances even when wetting and drying
takes place repeatedly.
6.3.2 Apparent Density and Porosity
Density is dependent almost entirely on the water to solid ratio or the
consistency of the plaster. Specifically, it is dependent upon the amount of
water that evaporates as excess water after setting, i.e., water that is not used
for hydration. Figure 6.6 shows that the apparent density has nearly a straight-
line relationship with the consistency ranging from about 1200 kg/ m3
共74.9 lb/ ft3兲 for 60 mL/ 100 g to about 850 kg/ m3 共53.1 lb/ ft3兲 for
100 mL/ 100 g.
Porosity is closely related to apparent density, that is, the pore volumes of
gypsum created by water in general lie between 50 % and 60 % depending on
the gauging water used. However, the effect of the configuration of the pores
and their structure can be quite significant. For example, hardened and dried
gypsum consists of two continuous phases, the solid phase consisting of
gypsum crystals and a gaseous phase consisting of air. This means that both the
solid phase and the gaseous pores are interconnected and for this reason the
structure of gypsum resembles that of a sponge rather than foam. Moreover,
the gaseous pores act as capillaries and this also has the effect that they absorb
and transport liquid water rather quickly.
Properties of Stuccos and Plasters 71

Fig. 6.7—Consistency versus apparent density and compressive strength.

6.3.3 Strength
It is perhaps not surprising that there is a close relationship between density
and the compressive strength of hardened and dried gypsum. The effect of
decreasing density on the compressive strength is relatively large 共Fig. 6.7兲.
Obviously both are caused by the water demand of the stucco, that is, its
consistency. The effect of the density on the flexural strength of gypsum is
similar but less well understood since flexural strength is a poor property of
gypsum.

6.3.4 Temperature Tolerance


A long-term exposure to temperatures above 46° C 共115° F兲 leads to strength
losses and must be avoided because gypsum converts into hemihydrate and
soluble anhydrite. It is often believed that the staying below 100° C 共212° F兲 is a
safe procedure. This misunderstanding about the calcination behavior of
gypsum and the difference between long and short-term exposure has led to
severe problems in practice, for example, with certain electrical heating
devices.

6.3.5 Fire Protection


Gypsum has been used for a long time to protect buildings from fire. A case in
point is the edict by Louis XIV stipulating that all wooden buildings must be
covered inside and out with a layer of gypsum. It is doubtful the mechanism of
fire protection by gypsum was fully understood at the time but considerable
protection of the building was certainly obtained. At the very least, it leads to
compartmentalization of the fire, which must be considered a tremendous
advantage. The more modern concepts of fire protection are referred to in the
next section.
References
关6-1兴 Kuntze, R. A., “The Aging of Gypsum Plaster,” Mater. Res. Stand., Vol. 7, 1967,
p. 350.
关6-2兴 Kuntze, R. A., ASTM Bulletin No. 246, 35–37, May 1960.
72 Gypsum

关6-3兴 Lane, M. K., “Disintegration of Plaster Particles in Water,” Rock Products, March
1968.
关6-4兴 Schiller, K. K., “Types of Plaster Hydration: Their Causes and Practical Signifi-
cance,” RILEM International Symposium on Calcium Sulfates and Derived Materi-
als, Saint-Remy-Les-Chevreuse, France, May 25–27, 1977.
关6-5兴 Ridge, M. J. and Beretka, J., “Calcium Sulfate Hemihydrate and Its Hydration,”
Rev. Pure Appl. Chem., Vol. 19, 1969, pp. 17–44.
MNL67-EB/Oct. 2009

7
Gypsum Board
GYPSUM BOARD IS UNDOUBTEDLY THE PREMIER PRODUCT MANU-
factured by gypsum companies in North America. In 1996, about 88 % of the
nearly 40 million tons of the total consumption of gypsum in the United States
was used for the manufacture of gypsum board. While approximately 3.3 mil-
lion tons were used as set regulator 共retarder兲 for Portland cement, very little
was used for the manufacture of other products such as plaster or gypsum fiber
board 共GFB兲. In Europe, plaster and GFB play a larger role.
A modern plant for the manufacture of gypsum board bears only a super-
ficial resemblance to the plant to manufacture board or lath patented by August
Sackett in 1894 关7-1兴. Following the usual pattern of development, the manu-
facturing process as well as the composition and structure of the board has
been reinvented on several occasions since that time. In fact, since its incep-
tion, the process is being constantly improved by the introduction of various
changes at most key points of the process. Among other things, the purpose of
these changes was to increase line speed, improve core to paper bond, and to
make the drying process more efficient. For this reason, the discussions of the
individual steps of the process emphasize the reason for the various functions
rather than describing fixed process requirements. The descriptions of the vari-
ous process steps, whether they refer to additives or mechanical devices, are
given only as examples.

7.1 Essential Manufacturing Features


The single most important factor in the manufacture of gypsum board is uni-
formity. This means uniformity of production rates, uniformity in feed rates
and raw material usages, and uniformity in stucco, water, and raw material
quality. The key to uniformity lies with satisfactory process control, equipment
maintenance, and adequate testing.
Stucco used for the manufacture of gypsum board has properties that are
similar to those of conventional stucco. However, there are three significant
differences with respect to both temperature and setting behavior. Stucco for
board manufacture is normally used at an elevated temperature between 100
and 120° F 共37 and 49° C兲 whereas conventional stucco remains at ambient
temperatures. Also, stucco in the board slurry is accelerated to less than 3 min
whereas conventional stucco is used with a stabilized setting time of 3 – 5 h.
The third significant point is the larger amount of gauging water used, that is,
about 90 mL/ 100 g whereas normal wall plaster requires about
55– 65 mL/ 100 g. Board stucco can be produced from natural rock or from
FGD gypsum or blends of both or with other synthetic gypsum. However, a
number of different demands are made on this material. That it can meet these
conditions shows again that stucco is a very accommodating substance. Fea-

73
Copyright © 2009 by ASTM International www.astm.org
74 Gypsum

tures such as weight, fire resistance, strength, water demand, and setting be-
havior are briefly discussed below.

7.1.1 Board Weight


When discussing gypsum board for North America, a variety of developmental
issues present themselves immediately. The most significant of these is the
replacement of the weight for regular 共non-type X兲 board by flexural strength.
This was introduced by the ASTM Committee C11 on Gypsum and Related
Building Materials and Systems, which constitutes a shift towards specific
physical performance requirements. This move was not followed by
jurisdictions in Europe and Japan where the weight replacement is apparently
not a significant issue.
Subsequently, technology was developed in North America to reduce the
weight of the board by maintaining compliance with physical performance
requirements, which were considered to reflect demands in practice. An
example of this situation is described in an article by R. Wenk 关7-2兴. In doing so
it was shown that most physical requirements could be met relatively easily
except nail pull resistance. In fact, nail pull resistance still remains the key
requirement that prevents an appreciable lowering of the weight beyond the
present level of about 1530– 1700 lb MSF 共7.2– 8.2 kg/ m2兲 for 1 / 2 in.
共12.7 mm兲 board.
7.1.2 Fire Resistance
Equally important, however, was the consideration that regular board with
reduced weight would not meet the fire-resistance requirements when tested
according to ASTM E119, which involves a specific test furnace producing a
standard heating curve 关Fig. 10.1兴. The temperature of the unexposed side of
the board when tested as an assembly using either load bearing wood studs or
non-loadbearing galvanized steel studs, must not rise above a certain
temperature after a described length of exposure time 关7-3兴. To meet these
requirements, a special fire resistant board “type X” was developed for each of
the board categories having a higher weight than standard board. This type of
board has a core that is usually modified by the addition of unexfoliated
vermiculite and glass fiber. Typically, a 1 h fire resistance-rating is required for
type X wallboard with a 5 / 8 in. 共15.9 mm兲 thickness and 3 / 4 h for board with
a thickness of 1 / 2 in. 共12.7 mm兲 关7-4兴. For type X lath, coreboard and
shaftliner board slightly different requirements apply apart from a higher
weight. The reproducibility of this test remains an issue since values of one
hour or less occur during the initial steep rise of the temperature curve.
It should be noted that assemblies using type X boards are listed in
manufacturers’ literature and by the Gypsum Association 关7-4兴. Where
applicable these listings also cover sound transmission classification 共STC兲 and
surface burning characteristics. Most of these assemblies are certified by
agencies such as Underwriters Laboratories. Random plant inspections are also
carried out to ensure adherence to validated formulations to establish, for
example, that type X board does not contain more than 2 % paper fiber
inadvertently or intentionally added to the core.
Gypsum Board 75

7.1.3 Board Strength


It should also be pointed out that while weight is a major factor in influencing
physical requirements, such as nail pull resistance, it is not the only one.
Various attempts have been described to increase the physical performance of
low weight board, usually by increasing the strength of its core. These attempts
generally involved a change in the structure of the core of the board by
producing larger foam bubbles and establishing layers with higher densities
near the core to paper interface. Similar attempts have been made by
increasing the amount of immobile starch that remains in the core 关7-5兴.
For example, it is claimed that additives such as sodium trimeta-phosphate
共STMP兲 increase the nail pull resistance of manufactured gypsum board and
having a beneficial effect on other properties as well. A patent by U.S. Gypsum
关7-6兴 states that STMP raises the compressive strength, rigidity, and
dimensional stability, such as drying shrinkage and nail pull resistance.
Rigidity in this context refers to humidified deflection, i.e., sag resistance in the
presence of high moisture levels of 90 %RH/90° F. Cubes cast in the laboratory
with accelerated slurry showed a strength increase of about 10 %–20 % for 0.1
% STMP additions. Manufactured board samples met the ASTM sag resistance
of 10/ 8 in. 共32 mm兲 with STMP additions of only 0.04 %. Drying shrinkage,
i.e., width and length changes on drying, was also eliminated completely by
STMP additions of 0.04 %. About 0.08 % STMP was required to raise the nail
pull from 88 to 96 lb for manufactured board. Perhaps somewhat surprisingly,
0.08 % STMP additions produced a slightly poorer bond failure. The addition
of both non-pregelatinized starch 共HI-Bond兲 and particularly pregelatinized
starch 共PF1000兲 was required to eliminate this problem, i.e., to obtain zero
bond failure.
In another patent by the U.S. Gypsum Company 关7-7兴, it is claimed that
tetrameta-phosphate also imparts increased resistance to permanent
deformation 共humidified deflection or sag兲 and dimensional stability 共drying
shrinkage兲 when added in quantities of about 0.04 %–0.16 %. This is
accomplished while maintaining or increasing strength. Specifically, it was
shown with laboratory specimens that the sag is less than 0.03 in. with 0.1 %
tetrameta-phosphate additions. It was also shown that there was no retardation
but there was little increase in strength. The claim that there is a considerable
improvement for drying shrinkage is questionable.
From a practical point of view it should be kept in mind that many factors
may have a profound effect on the size of the final gypsum crystals and thereby
the strength of the gypsum board. For example, a faster setting time produces
generally smaller crystals potentially producing higher strength. The amount of
water also affects the crystal size whereby short crystals and lower strength
may be obtained if insufficient liquid water is available.

7.1.4 Water Demand


An essential difference between normal wall plaster and the board slurry is an
increase in the water demand to increase its flow out of the mixer. The
intention is to obtain free-flowing slurries that spread evenly over the entire
76 Gypsum

Fig. 7.1—Snap set versus accelerated set.

paper surface upon which it is placed. This is particularly important when only
one main outlet is used for the mixer. An increase in the water demand of the
stucco shifts the yield point towards the origin and moves both density and
strength toward lower values. A similar effect is being produced by the addition
of surface active agents, such as dispersants and wetting agents.
7.1.5 Setting Behavior
Obviously, the setting behavior of the gypsum slurry is a most important
property and must be precisely controlled. The actual setting time will vary
from plant to plant depending on the combination of accelerator and retarders
used. To some extent, this is based on machine speed, the time of travel from
the board formation 共wet end兲 to the cut-off knife, and on the other additives
used.
The key property is the acceleration of the setting time of the slurry in the
time from the mixer outlet to the knife. In particular, short setting times in the
order of less than 3 min are achievable. Different accelerating techniques are
used but a combination of gypsum and potassium sulfate is common. The
gypsum provides homogeneous and heterogeneous nuclei and potassium
sulfate provides a large amount of sulfate ions immediately.
Schiller and Ridge have attempted to create formulas for the setting of
stucco that are based on sigmoid curves. However, as mentioned before neither
of these approaches captures the differences between the actual setting curves
which express different setting behaviors. For example, curves of temperature
rise versus time are given in Fig. 7.1 for a conventional accelerated setting
behavior and for a snap-set. Since the board slurry is normally used at a higher
than ambient temperature, this should be taken into consideration. The
differences between the two curves are important because they affect not only
the gypsum board manufacturing process but also the physical properties of
the board core such as strength. For the snap-set curve, the essential
Gypsum Board 77

Fig. 7.2—Measuring points.

differences consist of a longer induction period, a faster rate of rise, and a


sharp approach to a level that assumes complete hydration.
As shown in Fig. 7.2 the following points may be of importance when
analyzing setting behavior: 共1兲 Induction period 共the time interval without
measurable temperature development兲; 共2兲 angle of departure from baseline; 共3兲
time, duration, and fastest rate of temperature rise; 共4兲 angle of approach to
final degree of hydration; 共5兲 and the time for the final degree of hydration. The
final degree of hydration is often considered to be 95 % or 98 %. However, this
conflicts somewhat with the notion that a hydration above 95 % does not
translate into strength contribution to the board. It should also be kept in mind
that not all of these factors have to be determined at all times. However, it
should be established which factors are important for particular stuccos under
specific circumstances and these should be determined on a regular basis.

7.2 Core Manufacturing


The process of manufacturing the board consists essentially of six segments: 共1兲
Supplying the calcined gypsum containing additives to the mixer; 共2兲 providing
the paper liners to the forming table; 共3兲 addition of water and foam to produce
a slurry, which is then deposited onto the paper at the forming table, 共4兲 form-
ing the board on the forming table, 共5兲 conveying the formed board to the knife
to be cut and subsequently to the entrance to the dryer during which time the
hardening of the slurry takes place, and 共6兲 drying and subsequent handling of
the dry board.
The viability of the process to manufacture board requires very fast setting
rates. This is obviously the key function of the accelerator system, which con-
sists preferentially of ground and treated gypsum and sometimes of potassium
78 Gypsum

sulfate as well. Proper application of accelerators requires control over the


amount of gauging water and plaster rheology. In addition, it is necessary to
meet specifications of board properties, such as mechanical strength. Obvi-
ously, this requires accurate control of the structure of the gypsum formed as
well as of the amount and distribution of air voids in the core before hardening.
Although each segment plays an important role, the mixer and forming
table constitute the heart of the manufacturing process. The mixer is used to
prepare the slurry, which is then placed onto the back of a face paper liner
moving continuously over the forming table. At the end of the forming table,
the slurry is covered by a second paper liner forming the back of the board. The
edges of the face paper liner are then folded over twice forming an envelope.
After the stucco has hardened to a large extent, the boards are cut to the
specified length so that the ends of the board are the only part not covered by
the paper liner. The cut boards are then supplied to drying kilns where they are
subjected to a quick drying process. This method produces a sandwich struc-
ture whereby the paper provides the tensile strength and the core provides the
necessary compressive strength, hardness, and stiffness. A satisfactory bond
between paper and core is absolutely essential.
Additives are incorporated into the stucco to aid in the manufacturing pro-
cess, particularly with respect to water demand, air entrainment, rate of set-
ting, and bonding between paper liner and core gypsum. Weight reduction is
obtained by adding foam to the core. The water not required for hydration of
the plaster in the slurry is removed by a drying process, generally carried out by
kilns, which further reduces the weight of the board. It is essential to realize
that for the board making process the setting time of the slurry has to be
accelerated. Less than 3 min is the best time reached at present. Also, depend-
ing on the intended application of the board, materials are added to the core to
improve its physical properties, such as water absorption, flexural strength, or
brittleness. The paper is modified or laminated as well depending on the in-
tended application.
A schematic representation of the production arrangement is given in Fig.
7.3. The normal width of the boards produced with this arrangement is 4 ft
共122 cm兲, except when a smaller size such as lath is produced. The thickness of
the board is created by the forming plate or, less frequently, the master roll. The
length of the board is normally 7 or 8 ft 共213 or 244 cm兲 but other lengths can
be produced depending on limitations of the process. The thicknesses vary
from 1 / 4 to 1 in. 共6.4– 25 mm兲 depending on the specific type of board being
produced. Various edge configurations can be introduced as well 共Fig. 7.4兲. This
arrangement allows the manufacture of various types of board suitable for
specific applications in the field such as wallboard and predecorated wallboard,
backing board, coreboard and shaftliner board, soffit board, sheathing, base for
veneer plaster, and lath and ceiling board 关7-8兴. The board is intended for ap-
plication in the field as a panel without the use of water. For this reason, it is
often called drywall in reference to the main difference to the use of wet wall
plaster.
Gypsum Board 79

Fig. 7.3—共1兲 Face paper; 共2兲 forming table; 共3兲 back paper; 共4兲 conveyor belt; 共5兲 knife; 共6兲
board inverter; 共7兲 dryer inlet 共tipple兲; 共8兲 dryer; 共9兲 booker; 共10兲 transport; 共11兲 bundles;
共12兲 slurry preparation.

7.2.1 Stucco Supply


The initial step of board forming is the supply of satisfactory stucco to the
mixer. It is often not appreciated that the board making process can only be
optimized to produce a consistent quality if the stucco supplied to the board
mixer is uniform in terms of its physical and chemical properties. Quite often
the stucco handling system is responsible for introducing variability instead of
reducing it to a minimum.
The process of supplying stucco may be considered to begin with the stucco
supply bins, which are usually fed from the hot pit via a stucco supply screw.
Most new plants operate two stucco supply bins, which hold a total supply of
8 h for the board line. They are designed for mass flow and incorporate a high
capacity circulation system to continuously discharge, elevate, and blend the
stucco with the fresh stucco from the cooling system before returning it into

Fig. 7.4—Edge configuration of gypsum board.


80 Gypsum

the bins. This maintains a constant dynamic density, which is essential for
consistent stucco metering. The most common technique in new plants is to
introduce the hot stucco into a vertical airlift to directly flash cool and to
elevate it simultaneously. In some plants, a grinding mill is used as well.
Board core quality is enhanced with the correct slurry temperature and a
circulation system. This is intended to assure that the temperature of the stucco
is reduced to about 110– 125° C 共230– 257° F兲 and that the system is properly
vented to release excess water vapor. For this reason, most new plants are
equipped with a stucco cooling system ahead of the supply bins. This system is
considered ideal to produce the necessary uniformity of the stucco properties.
It is not recommended practice to use the stucco cooler to eliminate the supply
bin. Depending on the nature of the calcination process, it may be desirable to
keep the stucco briefly hot to achieve more complete calcination. Insulated
screw conveyors, hot pits, or short-term storage bins are often used for this
purpose.
7.2.2 Volumetric Feeder
The stucco is transported by a bucket conveyor from the supply bin to the
volumetric feeder. This feeder uses a variable speed drive to meter the correct
amount of stucco into the pin mixer. About 25 %–30 % of stucco fed to the
volumetric feed system is returned to the supply bin via another bucket
conveyor.
The mixing screw conveyor transporting the stucco to the pin mixer also
receives the dry additives from other conveyors. These materials are blended by
the mixing screw conveyor using cut and folding flights and mixing paddles.
They are then fed directly into the pin mixer of the board line.
If the board line is momentarily shut down, the dry mix conveyor is also
shut down and the metered stucco that normally feeds into the mixer is
returned to the recirculation system. This arrangement assures that fresh
stucco is immediately available for the mixer when the board line starts up
again. This is an important prerequisite for maintaining tight board weight
control and for high-speed board lines where clean start-ups are essential.
7.2.3 Additive Systems
In high-speed board plants, most wet and dry additives are received and stored
in bulk and material handling is kept to a minimum. Emphasis is placed on
feed rate accuracy and reliability. For example, stand-by pumps are usually
provided to allow maintenance flexibility and minimum down time. It is
important for the successful operation at high line speeds that very high quality
accelerator with uniform potency be precisely fed into the board slurry mixer.
Finely ground gypsum produced in water-cooled ball mills with an additive and
often being directly fed into the mixer is considered standard practice. This is
often augmented by potassium sulfate considered to be a chemical accelerator
to achieve the shortest set times.
7.2.4 Pin Mixer
Slurry mixing is technically the most critical aspect of the board plant,
especially at high line speed. A pin mixer is normally used for this purpose. It is
Gypsum Board 81

Fig. 7.5—Three-port mixer depositing slurry onto back of face paper. Note: slurries from
edge mixers.

produced in two sizes, 42 in. 共107 cm兲 and 60 in. 共152 cm兲. It mixes the stucco
plus additives that are received from the mixing screw conveyor with the
gauging water. A key problem is the build-up of freshly set gypsum inside the
mixer, which can be avoided by using a small amount of retarder such as
diethylene triamine pentaacetic acid 共DTPA兲. An essential feature of the pin
mixer is its role of incorporating foam into the water: Stucco slurry. The pin
mixer is often modified for a variety of reasons. For example, pins may be
replaced by feeding tubes or by plows. In some plants, mixers of different
designs or sizes are used. The main outlet from the mixer is a most important
feature and a single discharge is often utilized for this purpose instead of three
outlets shown in Fig. 7.5.
7.2.5 Roll Coater
Placing slurries with higher densities at the face paper interface is the
technique now generally used in the production of lightweight board. For this
purpose, separate mixers are provided to accept slurries from the pin mixer and
to remove foam from these slurries. They are then transferred to a paper roll
coater that places them onto the back of the face paper prior to depositing the
regular board slurry by a so-called foam placement technique. Tangential edge
mixers to reduce the quantity of foam from the slurry to be placed at the edges
of the board are no longer in use.
7.2.6 Forming Table
This is one of the most difficult operations of the board making process. Board
forming tables with top and bottom plates are often used. Control of the slurry
in front of the forming plates is a special problem due to the relatively violent
action at very fast line speeds. A forming plate or master roll is located at the
end of the forming table to control the initial thickness of the board. This
occurs when the back paper liner is placed onto the slurry that is already
spread onto the back of the face paper. The side edges of the face paper are then
folded over twice to form the edges of the board and creating an envelope for
the slurry. This process may be assisted by mud dams to contain the slurry as
82 Gypsum

TABLE 7.1—Approximate gypsum board formula


Weight
Ingredients Percent
Stucco 100
Water 94–98
Accelerator (HRA, heat resistant accelerator, sugar coated) 1.1–1.6
Starch (HI BOND, acid modification starch, 0.5–0.7
non-pregelatinized)
Dispersant (DILOFLO, napthalene sulfonate) 0.20–0.22
Paper fiber (hammer milled) 0.5–0.7
Set retarder (Versenex 80= DTPA) 0.07–0.09
Foaming agent (Witcolate 12760) 0.02–0.03
Recalcination inhibitor (Cerelose, dextrose) 0.13–0.14

well as by edging shoes in addition to the forming plate. The paper edges are
then glued to the back of the back paper liner after buffing to reduce thickness
and obtain a better paper to glue bond. The edges of the board are reduced in
thickness and modified to meet certain demands of application in practice.
Figure 7.5 shows an older type of mixer to illustrate the edge mixer
arrangements. Newer type of mixers have in most cases only one outlet, and
they may or may not have edge mixers.
7.3 Core Composition
To achieve this, the gypsum board manufacturing industry relies heavily on
chemical additives. They are intended to optimize the manufacturing process
共slurry workability, setting rate, and drying rate兲 and to ensure appropriate
physical properties of the hardened board 共density, strength, and water resis-
tance兲. Most gypsum board processes now typically include water reducers
共dispersants兲, foaming agents, and air entraining agents in addition to set modi-
fiers 共accelerators/retarders兲. While each additive is introduced for a specific
function, many of them will also influence the hydration process and each
other’s performance. Combination of additives can therefore interact in com-
plex and unpredictable ways to modify the setting process and the properties of
the resulting gypsum board.
Various slurry formulations exist depending upon the type of board to be
manufactured and various other requirements. A variety of components can be
combined to produce board with specific properties. Table 7.1 gives an example
of the ingredients used for the board formula. The relatively high amount of
water should be noted.
Table 7.2 gives an example of a board formula for a 1 / 2 in. 共12.7 mm兲 type
X board with a weight of 1700 lb MSF 共thousand square feet兲. The board is
produced at a relatively slow speed of 250 ft/ min which amounts to
1000 ft2 / min.
7.3.1 Gauging Water
The gauging water is obviously a key component of the board slurry.
Theoretically, it should meet all requirements for potable water. However, for
Gypsum Board 83

TABLE 7.2—Approximate formula for 1 / 2 in. type X board, 1900 lb MSF


Consumption per MSF Weight
Stucco 1745 lb 45 lb/ ft3
Starch 7 lb 30 lb/ ft3
Potassium Sulfate ⬃3 lb (if required) 90 lb/ ft3
Gypsum Accelerator 10 lb 95 lb/ ft3
Retarder (dry basis) 0.2 lb 95 lb/ ft3
Vermiculite 15 lb 55 lb/ ft3
Foaming Agent 0.05 gal 8.70 lb/ gal
Foam 300 gal 1.7 lb/ gal
Lignosulfonate 0.25 gal 9.85 lb/ gal
1 / 2 in. Glass Fiber 2.5 lb 95 lb/ ft3

economic or logistic reasons, other sources, such as deep wells or local streams,
may be used. In any case, care should be taken that they do not contain
components that are detrimental to board manufacture, such as sodium
chloride or sodium sulfate.
The water is usually heated so that the slurry is at 100° F 共38° C兲 or higher.
It should be noted that the amount of gauging water used is relatively high with
about 95mL/ 100 g. This exceeds the normal consistency of plaster which is in
the order of about 65mL/ 100 g.
7.3.2 Foam System
The objective of introducing foam into the board slurry is to create voids to
further reduce the final weight of the board without the use of water. Foam
addition also improves the physical properties of the board by making it less
brittle whereby a possible reduction in strength is minimal. It is also claimed
that foam improves the wet bond between core and paper of the board. In
contrast, large voids of collapsed foam are sometimes unintentionally
generated at the paper to core interface where this usually leads to strength
problems mainly because it interferes with the sandwich structure of the board.
The lack of stability of the foam is generally blamed for this situation. However,
it can be caused by a variety of other circumstances.
Foam is usually produced by a separate foam generator consisting, for
example, of either a disk type churn generator or a pump-type generator, which
mixes water containing a foaming agent with compressed air. The preferred
procedure is to dilute the foaming agent with water to about 5 %–10 % and
metering this solution to the foam generator. The feeding of the foaming agent
directly into the foam generator for a variety of reasons should be avoided.
In the past, foaming agents consisted of commercial soaps, such as rosin
soap, which characteristically produce dry foam. Dry or light foam is difficult
to feed homogeneously into the stucco slurry. However, soaps have been largely
replaced by synthetic material, which produces wet foams. These are known by
trade names such as Millifoam 共alkyl ether sulfate兲 or Cedepal 共ammonium
alkyl ethoxy sulfate兲. The foam is usually pumped directly into the pin mixer
84 Gypsum

where it is blended with the slurry. In some plants, the foam is combined with
the gauging water or alternatively the entire gauging water is fed through the
foam generator producing very wet foam. The stability of the foam generated in
a particular manner must be considered.
The production of foam with a consistent density and volume requires that
the make up water for foaming agent, and air be uniformly metered. A density
of 16 lb/ ft3 共256 kg/ m2兲 has been suggested but there are other densities in
use. This process requires that the metering pumps be accurately calibrated
and that information is available showing the consumption of each constituent
at various machine speeds. This is necessary because the drying characteristics
of the board would otherwise vary. However, the mix head should not be
controlled by varying the foam constituents other than in a minor manner.
In the past, it was considered that small and uniform bubbles uniformly
distributed throughout the core were required to obtain optimum strength of
the board. This view could be maintained as long as each bubble was
surrounded by core gypsum as in heavy board. However, this view had to be
changed as more foam was added to the core to reduce board weight further.
Bubbles began to touch each other and the core gypsum between them was
reduced significantly or disappeared at those points. In that way, the core
resembles a lightweight bone of a bird. For this reason, different theories have
been advanced recently regarding the size distribution of the foam bubbles and
their location to produce optimum properties of the board 共Fig. 7.6兲.
For example, gypsum board is now being produced that does not have
foam bubbles uniformly distributed throughout the core. In controlled density
共CD兲 board, the layers adjacent to the paper have fewer large bubbles and
consequently a higher density. To maintain the same board weight, larger or
more bubbles are being confined to the center of the core. When mounted on a
ceiling, this type of board has greater resistance to sag than regular gypsum
board. Other boards have a more sophisticated core structure in terms of
bubble distribution while maintaining a dense layer on one or both paper core
interfaces. The original work on this subject was carried out by Bob Bruce
共Nanaimo兲 et al. by demonstrating that lighter board with superior properties
can be obtained by controlling foam distribution in such a way that denser
layers occur at the gypsum core to paper interfaces. This is achieved by
selecting the stability of the foam to control the rate of collapse of the foam
which is caused by the rate of shear at the interface between the paper and the
gypsum core 关7-9兴. Subsequently, a similar effect was obtained by using blends
of two foaming agents that are chemically different and are classified as stable
and unstable 关7-10兴.
7.3.3 Accelerator System
The main purpose of the accelerator system is to shorten the setting time of the
board slurry from the mixer to the knife although it affects the entire setting
behavior of the board slurry. The time of set is usually the maximum point as
given by a time/temperature curve. Gillmore needles, thumb test, etc., are not
sufficiently accurate because they depend on the judgment of the operator.
Gypsum block accelerator is commonly composed of calcined gypsum that
Gypsum Board 85

Fig. 7.6—Micrographs of original and improved gypsum board core.

has been cast in 25 lb blocks and allowed to hydrate. For a less recommended
method, the gypsum block is obtained from the board slurry. The gypsum is
then ground just prior to board manufacture and added to the stucco slurry.
Immediate addition to the slurry is necessary because ground material, which
has been to be exposed to moisture for any appreciable time is noticeably less
efficient as an accelerator. In any case, the performance of the gypsum
accelerator should always be verified experimentally when it appears that it has
lost potency.
Another problem can arise with the block methods when the gypsum is
partially calcined during grinding because of the heat being generated. Partial
calcination reduces the acceleration capacity of the accelerator. One method to
solve this problem in practice involves an increase of the grinding speed to
produce more fine gypsum within a given time. This achieves exactly the
86 Gypsum

opposite effect since, under these conditions, the gypsum is dehydrated more
instead of less.
It should be kept in mind that increasing gypsum additions may decrease
the set time only to a certain point because the set is governed primarily by the
rate with which the stucco dissolves. This rate in turn is affected by such
factors as the temperature of the slurry, the fineness of the stucco, and the rate
of shear in the pin mixer. Therefore, additional usage of gypsum does not have
the desired increase of the acceleration effect when the fastest feasible setting
time of the core slurry is reached.
Gypsum accelerator is manufactured by passing ground gypsum or
landplaster through a ball mill and blending it with an additive. The most
modern acceleration systems involve ball milling in the presence of sugar
共USG兲, starch 共National Gypsum兲, or a suitable surface-active agent
共CertainTeed兲. USG produces HRA 共heat resistant accelerator兲 and CSA
共climate stable accelerator兲. The additives adhere to the surface of the gypsum
particles and in the case of sugar or starch prevents loss in acceleration potency
through partial calcination. In other cases, additives can also prevent aging,
that is, the loss of accelerating potency by moisture adsorption. Therefore,
these accelerators can be stored for a period of time or shipped in bags to be
added to the stucco as required. Protection from moisture is prudent even
under these circumstances.
The other common accelerator is potassium sulfate, which when added to
the board slurry quickly supplies sulfate ions 共SO−− 4 兲. It is considered a
chemical accelerator and as such is not affected by aging. It is generally used in
addition to the ground gypsum and it is considered to have a greater
accelerating effect. Ground gypsum has the tendency of creating initial
stiffening and a slow final set. In contrast, potassium sulfate does not promote
initial stiffening and it creates a snappy final set. Under certain circumstances,
potassium sulfate appears to form large gypsum crystals. This behavior is used
for sheathing type boards where it will often permit a reduction in asphalt/wax
emulsion.
Potassium sulfate is added to the stucco in the mixing screw or
alternatively it is added directly to the pin mixer. In some modern high speed
plants, the potassium sulfate is added just ahead of the single outlet port. It is
essential to provide means to immediately shut off the accelerator should the
board slurry stiffen prematurely. Attempts have been made to add potassium
sulfate to the pulp tank. However, this practice has the disadvantage that is not
possible to shut off the accelerator independent of the pulp.
7.3.4 Retarders
These materials are added to the board slurry to assure that no premature
setting occurs either in the mixer or on the forming table. Setting in the mixer
would lead to material build up that restricts flow and requires frequent
cleaning. Lumps may be formed in the slurry or a stiffening of the slurry may
occur, which produces cavities in the board core. These cavities appear often at
the core to paper interface and can result to poor bond to the paper. The use of
retarders also tends to minimize the initial setting action in the mixer caused
Gypsum Board 87

by minor quantities of gypsum that has not been calcined. It may also promote
correct accelerator use. Under these conditions, the retarder may enhance
slurry fluidity, provide for better incorporation of foam into the slurry, and
maintain a cleaner mixer. If optimum retarder usage is maintained, the wet
bond will also be increased. However, no amount of retarder will correct a
situation that is caused by severely under-calcined stucco.
In the past, dry retarders consisting of soluble keratin were added to the
mixing screw. Soluble keratin was obtained by hydrolyzing animal hoof or horn
material by treatment with sodium and calcium hydroxide. However, soluble
keratin has been largely replaced by sodium solutions of DTPA 关7-11兴. DTPA is
fed directly into the gauging water just ahead of the pin mixer and results in
better control of retarder addition. It also makes it possible to increase,
decrease, or shut off as required by the setting behavior of the slurry.
In board manufacture, a combination of retarders and accelerators are
typically used to produce a snap set 关Fig. 7.1兴. This type of set is characterized
by the lengthening of the induction period during where no noticeable
temperature is generated. This is followed by initial setting period and
subsequently by a more rapid setting as compared to the usual accelerated
setting. This also results in an earlier completion of the setting process.
Therefore, a snap set avoids setting in the mixer or on the forming table and it
assures complete hydration 共95 %–98 %兲 prior to cutting and drying. Also, snap
set maximizes the strength of the board core, which is particularly desirable for
the manufacture of lightweight board.
7.3.5 Starch
Starch is used in several industries primarily because of its high water-retaining
capacity. In the past, starch for lath or board manufacture was dissolved in
water by heating or cooking. However, modern board starches are modified to
be cold-water soluble and can therefore be added as a dry material to the stucco
in the mixing screw. A less desirable method of adding starch to the slurry is by
using the hydropulper of the pulp system, which results in less control.
The major sources of starch are wheat, corn, tapioca, and potato. In nature,
starch exists in the form of granules whereby their shape and size is determined
by the starch source. Chemically, starch is a long chain polymer consisting of
glucose units in the form of amylose as a straight chain and amylopectin as a
branched chain molecules. The amylose molecules consist of 300–600 glucose
units with an average molecular weight of 50,000–100,000. Amylopectin has a
molecular weight of over 1,000,000. The starch granules are formed through
hydrogen bonding.
Unmodified starch is insoluble in cold water. However, a process called
gelatinization starts at a temperature of 55– 60° C 共131– 140°F兲 when the
starch is heated in water. During this process, which continues until about
95° C 共203°F兲, the hydrogen bonds between the molecules begin to weaken and
the starch granules start to absorb water. This is indicated by an increase in
viscosity as measured by an Amylograph viscometer. Subsequently, the
granules start to swell until they are completely hydrated and they subsequently
collapse. As the result, the molecular chains are liberated, which is indicated by
88 Gypsum

a decrease in viscosity. Upon cooling, these molecules repolymerize by


retrogradation, which causes again an increase in viscosity. Despite this effect,
gelatinization is irreversible.
Physical modification leading to pregelatinized starches includes the
classification of granules into different sizes. This is done to increase the
solubility and the rate of migration during evaporation in the dryer, as well as
to reduce the cooked starch viscosity or to reduce the temperature of
gelatinization. It also involves drum drying of pregelatinized starches,
extrusion, and the formation of expanded foams. Chemical modification which
produces non-pregelatinized starches includes the reduction of the molecular
weight of starch by acid thinning with H2SO4 or HCl or oxidation by forming
carboxyl carbonyl groups. It also includes esterification, etherification, cross-
linking, as well as the formation of copolymers.
Various hypotheses have been advanced in support of concepts to explain
the bond between gypsum crystals and paper fiber. To a large extent, these
hypotheses are based on the fact that in the past gypsum board was produced
with slow drying and without the use of starch. However, it was found
subsequently that pre-gelatinized 共cooked兲 corn starch allowed much higher
drying temperatures without breaking the paper to core bond. This was the
discovery that allowed the high speed manufacture of modern gypsum board.
Despite these efforts, the question of how starch bonds the core to the
paper remains largely unanswered. The most common of these hypotheses
assumes that when the board enters the dryer the core begins to heat up and
the starch 共or a fraction of it兲 will dissolve in the free water of the core slurry. As
the board dries, both the water and the dissolved starch move towards the
surface paper of the board. The starch then precipitates at the point where the
water turns into steam at or near the core/paper interface. It forms a gel at this
location covering the gypsum crystals. Since this gel has a high water holding
power, it will maintain liquid water around the gypsum crystals protecting
them from calcination. This assumes that the bond between core and paper is
formed by gypsum crystals interlocking with the paper fibers of the inner ply of
the liner with the starch playing a protective role.
A different hypothesis attempts to take into account the performance of
new types of paper. It suggests that starch alone bonds the gypsum core to the
paper. The starch produces the bond because modern papers have lower water
absorption than older paper and therefore absorb less water that is
supersaturated with respect to gypsum from the board slurry. For this reason, it
is unlikely that gypsum crystals form within the paper to create a bond. For
this, it is visualized that the starch bridges the paper fibers with the gypsum
crystals.
Regardless of which hypothesis is correct, starch is essential for obtaining a
good bond between core and paper particularly when board is produced at high
speeds and high dryer temperatures. Good starch migration profiles exist when
75 %–85 % is found within the paper and at the core/paper interface. About
65 %–75 % of this quantity is obtained through migration from the center of
the core. This migration depends to a large extent on the degree of
Gypsum Board 89

modification, which increases the mobility of the starch as well as the cold and
hot water solubilities. For example, a migration of 45 % is obtained at a
Brookfield Viscosity of 900 cps 共12 % solids, 80° C兲 whereas the migration is 80
% at a viscosity of 20 cps.
Other factors influencing starch migration are the water to stucco ratio of
the board slurry, that is, the free water available to act as the carrier for the
starch. For example, a migration of 70 % is obtained at a ratio of water: stucco
of 82 mL/ 100 g, whereas the migration is only 30 % at a ratio of 92 mL/ 100 g.
For a similar reason, an increase in foam voids reduces starch migration.
Starch migration is also affected by the sizing and porosity of the paper.
Starch will migrate following the path of least resistance thereby accumulating
greater amounts at the more porous interface. The face and bottom papers
should have similar properties in order to accumulate equal amounts of starch.
Likewise, the drying process is often variable from deck to deck and from left to
right. In general, a faster rate of water evaporation results a lower migration
level. Similarly, different drying rates will cause disproportionate starch
migration to the interfaces.
Starch is particularly important when chlorides are present in the board
slurry that can interfere with the bond to the paper. This is presumably due to
the lowering of the calcination temperature of the gypsum during drying.
Because starch raises the calcination temperature, it can offset the negative
effect of the chlorides.
It has also been found that due to its molecular weight distribution, a
certain amount of starch remains within the core providing adhesion between
gypsum crystals. This results in an increase of the compressive strength of the
core. However, the quantity of starch used to produce this effect considerably
exceeds the normal amount of starch used. Increased quantities of starch can
lead to various problems including long drying time because of the water
holding capacity of the starch.
In summary, the principal function of the modern starches is to protect or
create the bond and to a degree to increase the strength of the board. For this
reason, starch is often increased as a cure-all for bond problems. However, this
may require an increase in the drying temperature because the board becomes
harder to dry, which usually compounds the problem. Obviously, the stucco
quality should be improved and maintained so that the lowest amount of starch
can be used instead of fixing a problem with an expensive additive.
7.3.6 Other Additives

7.3.6.1 Boric Acid


Boric acid has been used in the manufacture of gypsum board for about
60 years to increase sag resistance particularly of ceiling board, which sags
both under its own weight and the weight of added insulation. In addition,
boric acid has been used to prevent paper bond failure caused by calcination in
the dryer. The manufacture of lighter board has led to an increase in the use of
boric acid, as sag and calcination on overdrying become more problematic.
Other uses have been for increases in flexural strength, edge and core hardness,
90 Gypsum

and nail pull resistance. Most manufacturers use boric acid at levels between
0.06 % and 0.3 %. If the level of boric acid addition is much higher, the board is
more brittle and stiffer. Consequently, the breaking load is concentrated over a
short span at the center of the board when the board is lifted on one end during
handling, which results in easier cracking and breaking.
Examinations of the effects of boric acid have shown that it decreases the
rate of calcination substantially but not the onset of the calcination. This effect
on the rate of calcination is indicated by the increasing peak temperatures of
differential scanning calorimeter curves of set stucco to which increasing
amounts of boric acid has been added. This should not be confused with a shift
of the onset of calcination, which remains the same as shown by the incipient
temperatures of the DSC curves. The protective effect of boric acid on the
calcination of the board in the dryer must therefore be considered in terms of
time as well as temperature.
It has also been shown that boric acid migrates easily and rapidly to the
surface of the board. It was further shown that it appears to modify the crystal
habit at the surface of the board but not within the interior of the core. There is
no explanation for this effect since boric acid migrates to the surface after the
plaster has set.
As expected, boric acid increases the humidified strength of the bond
between paper and core and it does so at a lower concentration than starch.
There appears to be a beneficial effect in the presence of chlorides. Also, boric
acid has a noticeable effect in improving the humidified deflection 共sag兲 of the
board.
Boric acid improves the edge hardness and nail pull resistance of board
particularly in the presence of chloride concentrations that are above
specifications 共e.g., 200 ppm兲. For unknown reason, this effect seems to be
dependent upon the origin of the stucco used.

7.3.6.2 Water Reducing Agents


Dispersing agents or wetting agents are often added to the board slurry to
reduce the water requirement of the stucco or to increase the fluidity of the
stucco slurry maintaining the water solid ratio. The use of these agents is
considered a valuable tool in reducing the quantity of water used for the slurry
because it results in faster machine speeds and reduced manufacturing costs.
At the same time, it is recognized that these materials may change the yield
point or the other rheological properties of the slurry. It is estimated that a
reduction of 1 % in gauging water can lead to an energy saving of approxi-
mately 1 % through reduced drying requirements. Obviously, the water saved
had to be replaced with foam.
The most common of the dispersion agents are calcium, ammonium, or
potassium lignosulfonate 共Lignosol兲. Most are available as solids and they can
be added to the stucco in the mixing screw or as both solid and liquids to the
pulp tank. Care should be exercised when using these materials since some
have retarding properties and may affect the gypsum crystal formation and
lead to a weaker core or a poorer core-to-paper bond.
Gypsum Board 91

More modern materials or superplasticizers, such as sodium salt of


polynaphthalene sulfonic acid 共PNS兲, polymelamine sulfonates, or
polyacrylates are used in place or in addition to lingnosulfonates. They are
available in liquid form and hence are added directly to the pin mixer or the
gauging water. Addition through the pulp tank is possible but the previously
mentioned loss of control can occur. In general, these agents are water-soluble
organic polymers that have ionized groups as well as hydrophobic groups 共side
chains兲 and a molecular weight from 103 to 105. It has been stated that PNS at
a concentration of 0.1 % by weight of the stucco has the same effect as
increasing the water by 10 %. This is based on a similar increase in patty size
when water or PNS are added to a stucco paste with a normal patty size 关7-12兴.
A variety of theories have been advanced to explain the mechanism of water
reduction by dispersion or wetting, such as adsorption, electrostatic repulsion,
or steric repulsion. In the presence of calcium ions on the surface of the stucco
surfaces, the most likely of these is adsorption. However, it should be noted that
electrostatic and steric repulsion could seldom be eliminated entirely. Also,
superplasticizers have side effects in many cases such as lengthening the setting
time and modifying crystal growth and crystal morphology. It is also significant
that they may interfere with or have synergistic effects on other additives such
as foaming agents.

7.3.6.3 Paper Pulp


Paper fiber is considered to improve some of the physical properties of the final
board, particularly brittleness and nailability. For special fire rated type X
board,
2 % by weight must not be exceeded to meet UL inspection. In this case, paper
fiber is usually introduced through recycled board. For regular board, however,
paper pulp is produced by dispersing used newsprint or waste board paper in a
hydropulper on a batch basis for about 60 min. The hydropulper or similar
equipment pulps the paper under high shear, forming a homogeneous
suspension. The aqueous pulp slurry is then transferred to a holding tank
共chest兲 holding about two to three batches. This tank is fitted with a stirring
mechanism to prevent settling of the paper. From this tank, the pulp
suspension is pumped continuously into the pin mixer or the gauging water
line if it is free of foam. In some plants, this suspension may also include
retarder, core starch, potassium sulfate, and water reducing agents. Obviously,
a much preferred procedure would be to meter the materials separately into the
pin mixer.
Usually the pulp tank supplies the major part of the gauging water but in
some plants only a small volume of water is incorporated with the dispersion.
In this case the remainder of the gauging water required for the slurry mix is
supplied by a separate water source. This is a preferred procedure since the
gauging water is metered directly into the pin mixer.
92 Gypsum

7.3.6.4 Hydrophobes „Asphalt Wax Emulsions/Silicones/Siloxanes…


It is a definite characteristic of the core of untreated gypsum board that water
is adsorbed by positive capillary action. This is due the fact that the core of
gypsum board consists of a continuous solid phase 共gypsum兲 and a similar
continuous phase existing of empty space. Capillaries are formed in this space
as the water evaporates, that is not needed for hydration, and these capillaries
are clad by gypsum. This can be demonstrated by placing the board
horizontally in contact with water whereby the board will show a wicking
effect, that is, the board becomes wetted above the water level.
Consequently, the purpose of hydrophobes is to create a negative capillary
action 共negative meniscus兲 to avoid penetration of water into the core of the
board for a predisposed time. ASTM C473-07,1 Section 20, addresses this
situation by specifying that the Water Resistance of Core-Treated Water-
Repellant Gypsum Panel Products be determined by submerging a specimen of
board in water for 2 h with a 1 in. hydrostatic head.
It is an accepted hypothesis that asphalt/wax emulsion functions by forming
a coating over the gypsum crystals, which surround the capillaries within the
core structure. Microcrystalline wax crystallizes on the surface of the asphalt in
the dryer. This creates a negative meniscus for these capillaries and prevents
the water from entering the core for a certain period of time. There is a trend
towards replacing the asphalt portion in the emulsion by wax that contains a
non-microcrystalline portion, which improves its handling properties but
otherwise achieves the same effect.
In the production of water resistant core for interior backing board 共base for
tiles, etc.兲 or exterior sheathing, asphalt wax emulsions are added to the board
slurry. An example of the consumption of asphalt/wax for a 1 / 2 in. 共12.7 mm兲
board is given as 13.2 gallon 共50 L兲. The ratio of asphalt to wax ranges from 2:1
to 4:1. The paper of this type of board is also water repellant.
The asphalt/wax emulsion is generally supplied in heated tank cars and is
often fed from these cars into the pin mixer using heated lines. Alternatively,
heated tanks are available at the plant for storage. Attempts to add these
emulsions directly to the pulp tank have led to problems. Avoiding this practice
would make changeover easier and avoid foaming of the pulp system. Needless
to say, all pump, piping, and related equipment should be flushed with water
after completion for the board run so that fouling can be avoided.
The total emulsion usage and the quality of the emulsion largely determine
the water resistance obtained. However, there are other conditions that
influence this property. For example, water resistance decreases with the
amount of pulp fiber added. Also, water resistance depends on the amount of
foaming agent used whose efficacy is reduced by most emulsions. Finally, water
adsorption is decreased by the accelerators used. This has led to an increase of
potassium sulfate as compared with ground gypsum.
Apart from water resistance, asphalt/wax emulsions or wax emulsions alone
do not improve the other physical properties of the board and in fact they may
1
Standard Test Method for Physical Testing of Gypsum Panel Products.
Gypsum Board 93

be detrimental. In particular, core treated boards have poorer sag resistance


than regular board. Initially it was suspected that this was due to the presence
of the organic material. It is now believed that this is due to larger amounts of
potassium sulfate being added to the board slurry to obtain the required setting
time. Potassium sulfate produces large gypsum crystals, which may be the
reason for the poorer sag resistance of the board.
Most of these problems can be avoided when silicones or siloxanes are used
instead of asphalt/wax emulsions. These materials were first used in Europe,
where heavier board is manufactured. For lighter board in North America, this
approach was initially not very successful. However, these problems have been
overcome as indicated by the steadily increased use of these materials for the
purpose of manufacturing water repellant board.
Silicones or siloxanes create a negative meniscus; the requisite by which
they are effective in increasing water repellency. They apparently do so to by
adhering to the gypsum forming the inside of the capillaries. However, this is a
questionable proposition and it raises the question of the adhesion of the
asphalt/wax emulsion to the gypsum crystals, which so far was thought to be a
necessity. Therefore, the asphalt may act merely as a diluents and the
hypothesis of asphalt adhesion to the gypsum crystals should be revisited.

7.3.6.5 Glass Fiber, Vermiculite, and Clay


Glass fiber is added to the core of special fire rated type X board to change the
crack formation of the board when it is subjected to fire. This results in a larger
number of small cracks instead of fewer larger ones. For the same reason, glass
fibers improve the mechanical strength of the board when the gypsum has been
calcined on exposure to fire.
Increasingly, there is a tendency to replace pulp with glass fiber to increase
the flexural strength of the board. This avoids the breaking of long board
共10– 16 ft or 3 – 4.9 m兲 that is often lifted in the field at one end. Glass fiber can
be supplied to the plant as dry strands. These fibers are chopped to 1 / 2 in.
共12.7 mm兲 and fed into the dry stucco in the mixing screw. This is not a
recommended procedure because it is difficult to mix chopped fiber with dry
stucco. A more suitable procedure is to add the glass fiber to a separate tank. In
some plants, this is usually done for pre-chopped glass fiber that is treated with
a wetting agent to aid dispersion in water.
Unexfoliated vermiculite is added to the board slurry when special fire rated
type X board is produced. It is generally metered into the stucco in the mixing
screw. Vermiculite ore exfoliates, that is, it expands resembling an open
accordion when subjected to heat. On exposure to fire, this expansion
compensates for the shrinkage of the core due to the calcination of the gypsum.
As a result, the cracks and fissures formed in the core are fewer and smaller,
which increase the fire resistance of the board. Additionally, the exfoliation
process absorbs energy, which provides further protection from heat transfer
through the board.
When special fire rated type X board is produced, some plants add clay to
the board slurry in addition to vermiculite and glass fiber. Clay is generally
94 Gypsum

added to the stucco in the mixing screw. The exact mechanism for the function
of this material is not known but it increases the fire resistance of the board.
However, clay should not be used unless absolutely necessary because of the
negative effect on the setting characteristics and the water demand of the
slurry.

7.4 Paper Liners


7.4.1 Paper Handling
For board manufacture two roll arrangements are available, one each for the
face and the back paper of the board. The face paper is laid down first and face
down on the forming table to receive the slurry on its back. The back paper is
then placed on top of the slurry forming the top of the board as manufactured.
Each of these paper roll arrangements consists of the following:
1. A paper roll stand.
2. Pull rolls to protect the paper from breaking.
3. Guides and tensioners.
4. Paper heaters to remove absorbed moisture.
5. Units to cut grooves along the inside of the face paper for easier folding.
6. Paper cutter if board with smaller width is to be manufactured.
Shaftless roll stands and automatic splicers are the standard equipment for
new high-speed board lines. The roll stands are capable of handling roll
diameters up to 84 in. 共2.1 m兲 and all normal paper widths. Handling of rolls
by fork trucks and manual handling is minimal. A system is generally provided
that incorporates all controls for accurate positioning of the paper rolls,
automatic splicing, and continuous automatic tension control. Splicing must
be very accurate and reliable. Special features provide power assisted
unwinding to facilitate paper thread-up and smooth start-ups.
These systems are relatively easy to operate and require minimum operator
attention. The only other devices used are gas-fired heaters to condition the
paper as well as web guides to precisely track the paper sheets into the forming
station and creasers for folding.
The stand supplying the face paper is usually located immediately adjacent
to the board forming table. This table provides support for the pin mixer and
the vibratory roll. The vibratory roll, not used by all plants, is intended to
assure that the slurry discharged from the pin mixer moves across the width of
the paper.

7.4.2 Paper Liner Performance


Some plants still experience problems that should have disappeared with
modern developments. For example, difficulties may be encountered because
the paper shrinks and expands as it absorbs moisture. Most commonly, the
edges of the paper expand when stored at a high relative humidity resulting in
cockles near the edge of the board during manufacture. For this reason, most
plants heat the edges of the paper to avoid this situation.
Once the board is formed, there are several factors that affect the
performance of the paper. For example, gypsum board may be manufactured
Gypsum Board 95

with face and back papers having different expansion characteristics. The
expansion of paper will vary with the quantity of waste paper used, the
saturation, the permeability, and the moisture of the paper. For this reason, it is
not safe to estimate the expansion by saturation or permeability alone.
The wet bond as observed at the knife is another consideration. This bond
is good when a fairly thick layer of paper adheres to the core. Again, there are
many factors that influence this behavior, such as the temperature of the slurry,
the type and amount of foaming agent, the saturation of the paper, the water to
stucco ratio of the slurry, the setting time characteristics, and the amount of
retarder/accelerator used. Surprisingly, it is not affected by the type and
amount of starch in the core.
As the board enters the dryer, the permeability of the paper becomes a
critical factor. If the paper has too low a permeability, entrapped steam will
form under the paper. This may result in formation of blows or blisters when
the steam pressure exceeds the wet ply strength. Conversely, if the permeability
is too high, the board may dry too easily and peelers may result because of
surface calcination of the board. For this reason, it is extremely important that
face and back papers are balanced in terms of permeability as close as possible.
In recent years, the quality of the paper used has improved so that most of
the problems experienced in the past have disappeared. For regular board,
ivory or cream paper is used for the face whereas grey paper is used for the
back of the board. For some face papers, only the top ply consists of cream or
ivory material and the remaining plies are made up of gray stock. Ivory paper
stock is processed to remove impurities. The total weight of the face and back
paper is about 90 lb per MSF 共0.5 kg/ m2兲. Water repellant paper used for
sheathing is colored brown usually because of asphalt impregnation. Different
producers use different colors for other types of board. For example, USG uses
green water repellant paper for interior backing board whereas the National
Gypsum Co. uses blue paper.
The main improvements are due to the use of fewer plies and the use of
waste paper. These developments occurred despite a decrease in grammage
from 380 GSM in 1965 to 190 GSM in 1996. Over the same time, the machine
speed was increased from 110 to 650 m / min, which resulted in an increase of
the production rate from 7 to 21 tph.
Modern papers consist of three to seven plies whether they are intended for
the face or the back of the board. For the face paper, the first ply that forms the
ivory exterior of the face is normally decorative. The grey back ply 共normally
one or two兲 is intended to bond to the board slurry. The center of the paper is
formed from one to four middle plies. All these plies are held together by
hydrogen bonding.
A typical paper liner consists of 87 % by weight of cellulose 共paper兲 fibers,
10 % water to control paper expansion and paper handling, and 3 % chemicals
of various types. These chemicals are intended to provide water repellency
共sizing兲, color, anti-scuff properties 共starch兲, anti-mold properties 共fungicide兲,
and additional strength 共starch兲. For economic manufacture, the paper liner
96 Gypsum

must have a low grammage to allow high productivity and consistent low cost
of raw material.
The paper is produced by various machines including underfelt formers,
ultra-formers, Manchester formers, and high-speed, multi-wire Fourdriniers. A
typical papermaking machine has a 400,000 ton capacity and operates at
speeds of up to 700 mpm. This equipment produces trimmed width of either
2.85 m 共9.35 ft兲, 3.65 m 共11.97 ft兲, or 4.44 m 共14.56 ft兲.
To assure that a good quality gypsum board can be manufactured, the
paper liners must meet the following requirements at the board plant: A good
reel presentation 共correct width, diameter, etc.兲, true and consistent unwinding
共no slack edges, nicks, holes, or creases兲, ease of scoring and producing good
edges, acceptable gluing properties at edges, predictable expansion in CD
direction 共no cockling or ridging兲, good wet bond at knife, resistance to water
ingress to retain strength at turn over, easy to dry in board driers, excellent dry
bond, and good scuff resistance.
In addition, the following properties are an issue in the marketplace:
Presentation of attractive surface 共flat surface, no creases, color, etc.兲, ease of
decoration by paint or wallpaper, acceptance of joint compounds or joint
cements and wallpaper, and good strength 共handling without breaking兲.

7.5 Conveyor Section to Dryer


7.5.1 Board Conveyors
After the slurry is formed on the forming table and encased by the paper, the
board moves automatically to a belt and subsequently to live roll sections. At
this stage, the part of the face paper overlapping the back paper is often
sprayed with water to improve the glue bond. The belt section generally
consists of two belts that are supported by a series of precision rollers on a steel
framework. Each belt is equipped with pulley terminals, AC motor drive for
firm control in high speed plants, and belt washers. Their function is to
maintain the board in flat position while it is moved from the mixer towards
the knife. The live roll section consists of a series of AC powered stainless steel
rolls that transfer the board, now sufficiently rigid, from the belt section to the
knife. In newer plants, forming belt conveyors reach almost all the way to the
knife. This is more economical than open roll sections and provides a better
support for the gypsum board.

7.5.2 Knife
The cut-off knife with a digital control system is used to cut the board to
lengths usually ranging from 4 to 16 ft 共1.2– 4.9 m兲. The requirements for
accurate cutting of the wet board at very fast line speeds led to the development
of a new rotary cut-off knife. The new design uses separate drives on the top
and bottom shafts, which are synchronized electronically. Because of the low
inertia design, the knife reaches an exact speed and position to produce a clean
cut and precise board length. There may also be a feature to perforate the
board at a specific length for breaking at the job site.
It is essential that the length of time from the mixer to the knife is sufficient
Gypsum Board 97

for the board to develop the required strength and stiffness. This means that the
stucco in the core must have hydrated sufficiently prior to cutting. A typical
setting time of board slurry for new boardlines provides about 3 min between
the mixer and the cut-off knife. Many older plants have increased the speed of
manufacture and operate in the 3 – 4 min range. On the line, the degree of
setting is generally tested by a thumb impression. This underscores the
importance of the snap set, which allows a better control of the setting process.
A serious flaw of the board plant design is that the boardline is often too short
to allow speed increases in the manufacturing process. As a result, problems
occur either at the mixer when the slurry sets too early or at the knife when the
slurry has not hardened sufficiently.
The time required from the knife to the dryer is approximately another
4 min. Flexing of the board during that time period must be avoided. It is
important that the stucco in the core of the board has fully hydrated when
entering the dryer. This is often overlooked and it is assumed that sufficient
strength at the knife guarantees complete hydration at the dryer entrance,
which is not always the case. Shortly after entering the dryer, hydration of the
stucco ceases because of exposure to high temperatures.

7.5.3 Acceleration and Transfer Section


Some high-speed plants designs are based upon cutting long wet boards and
then sawing them to the desired final length after drying. The cut lengths of the
board are 24 ft 共7.3 m兲 minimum and 32 ft 共9.8 m兲 maximum. They are fed
two pieces at a time into the acceleration section. The purpose of this section is
to quickly move the board away from the knife. In the transfer section, the
boards are inverted so that they pass through the dryer in a face-up position to
avoid staining of the face paper. Before entering the transfer section, the boards
are inspected visually and any board showing sign of flaws or damage is
rejected at the reject conveyor stage.
The boards are “degrouped” before feeding them into the inlet section of
the dryer, which reduces the required length of this section. The overall length
of the inlet section is usually selected to obtain a total of about 7 min of
hydration time from the mixer to the dryer entrance. The inlet section is
entirely belt driven for smooth, quiet operation and low maintenance. Fast and
smooth motion as well as accurate positioning is achieved by special servo
drives for the dryer tipple.

7.5.4 Bond between Core and Paper


The dryer and the drying process cannot improve board quality but it can have
a negative effect. For example, improper drying may cause poor bond of the
paper to the core. Two types of failures occur most often under these
circumstances, commonly known as splitters or peelers. Splitters are
distinguished from peelers by having some of the core material attached to the
paper. In contrast, peelers are characterized by a clean paper surface with
almost no core material attached. Although there are numerous hypotheses for
the occurrence of splitters and peelers, their mechanism is not clearly
understood. However, there seems to be a positive correlation between chloride
98 Gypsum

content and paper-to-core bond failures. Other types of failure occur less
frequently. For example “blows” refer to the lack of paper core bond over large
areas of the board.
A number of hypotheses exist for the generation of a good or poor bond
between gypsum crystals and the paper fibers of the liner. While some of these
make sense, there is little evidence to support them. For example, it is claimed
that the gypsum crystals interlock with the paper fibers protruding from the
paper liners, thus bonding the paper firmly to the core by a mechanical bond as
opposed to a chemical bond. This overlooks the role of the starch functioning
under modern conditions 共see Section 7.3.5兲.
Small gypsum crystals tend to form under unsatisfactory circumstances
within the pin mixer. These will be broken by mechanical action and do not
continue to grow and in severe cases produce a weak paper-to-core bond. In
extreme cases it can also produce a soft core. Similarly, it is believed that if the
crystals attain any appreciable length before the slurry contacts the board
surface liner, the bond will be weak because the crystals are easily broken by
mechanical action. More believable is the assumption that the strongest paper-
to-core bond is obtained when the natural crystal growth does not begin until
after the board has been formed. If lumps of set gypsum enter into the core,
there will be no bond in the corresponding areas and this can lead to blisters
and blows.
Other process factors such as accelerator-retarder combinations should be
investigated as well. Apart from bond considerations, a slurry of stucco, which
stiffens too quickly can result in blisters of the paper and splitters can occur on
the face or back of the board. If stiffening is due to excess accelerator, side
effects may include excessive shrinkage of the board during drying, as well as a
“punky” hard drying core.

7.6 Board Drying and Handling


7.6.1 Board Drying
It is essential to control this process so that the gypsum slurry is not calcined
again due to the high temperature in the dryer, which would lead to loss of core
strength and loss of paper-to-core bond. Initially, the board is moved into the
dryer by a feed tipple, a conveyor that is hinged on one end to allow vertical
movement and access to all dryer deck levels. Most new plants in North
America are built with board dryers that provide 35– 40 min of drying time for
1 / 2 in. 共12.7 mm兲 thick board. This is somewhat conservative considering that
the typical drying time in older plants is 30 min or less. Board dryers are
available that function in a longitudinal, a cross-flow format, or a combination
of the two.
The dryers are either two or three boards wide. The wider type is
recommended because they are more economical. They also provide significant
timing advantage for the dryer feed system as well as the dry board handling
equipment, which is an important benefit for high-speed plants. Most new
dryers are 12 decks high, which reduces their overall length as compared to the
typical eight deck dryers used previously. Each deck is individually driven and
Gypsum Board 99

speed controlled. The fans are also speed controlled to adjust air volume and
velocities.
Three zones are common but the largest dryers have four zones. For three
zone dryers, the first zone operates at the highest temperature usually at about
600°F 共315° C兲; the second zone operates at about 400– 450°F 共205– 232° C兲;
and the third zone at about 350°F 共178° C兲. For each zone, the inlet and outlet
temperature of the hot gases is monitored. Typically, the first and second zones
remove each about 40 % of the excess water. The board experiences a rapid rise
in temperature in the first zone and a corresponding decline in the second zone.
Removal of the remaining 20 % excess water in the third zone is carried out at
the lowest temperature and it is therefore a somewhat slower process.
Most dryers are heated with natural gas, but oil heated and to a lesser
extent coal heated dryers are in use. Standby fuel may be propane or light oil.
The dryers are very energy efficient and some of the new dryers are therefore
not equipped with heat exchangers. However, by optimizing individual zone
exhausts the energy requirement can be improved from ⬎1550 BTU/ lb
共 ⬎ 780 kcal/ kg兲 to 1400 BTU/ lb 共710 kcal/ kg兲. A further improvement can be
obtained by exhaust heat recovery, which reduces the energy requirement to
⬍1300 BTU/ lb 共 ⬍ 650 kcal/kg兲. For maximum efficiency, some dryers are
analyzed in detail relying on real time balancing tools, automated dryer firing,
etc.
To avoid calcination of the ends of the board, it is essential that they move
through the dryer in close contact. Both the ends and edges of the board are
subject to calcination when the temperature of the dryer is too high or the
when the speed through the dryer is too slow.
7.6.2 Dry Board Handling
The take-off system is intended to be capable of dealing with all board width,
length, and thicknesses at full production speeds. It should handle board that
has the same finished lengths or two or three different lengths at the same time.
After the take-off procedure, the boards are separated by another accelerator
section. A subsequent transfer section moves the board at a 90° angle to the
booker that places two boards face to face. Two separate systems are provided
for transferring, booking, and sawing, each capable of processing the full flow
of board. One system is intended for board rejects or length change transition
and as stand-by.
The ends of the boards are subsequently cut to obtain the desired length.
Long board are squared and clamped during sawing with traveling saws at slow
traveling speed. This results in a cut of high quality and accuracy.
After booking, the boards are accumulated into stacks up to 4 in. 共100 mm兲
high, which greatly reduce the time required for the downstream functions.
Each bundler is equipped with an accumulator to collect two or three bundles
before feeding them to the stackers. Two separate stackers are supplied for each
bundler, which operate alternately and are capable of building multiple stacks
to minimize forklift-handling time.
In North America, all board is fitted with end tapes. Since no sawing is
being performed at the bundlers, a faster throughput can be achieved. In a
100 Gypsum

separate operation, special boards are finished with vinyl sheet, paint, or fitted
with aluminum foil to their back. Also, dunnage or spacer strips used to
separate stacks of board are produced in a separate operation. Several reject
stackers are provided to deal with problems as they arise. Reject board is
sometimes fed back into the system during full production operation.

7.7 Joint Compounds and Cements


Proprietary joint compounds 共drying types兲 and joint cements 共setting types兲
are used to fill and finish the joints between the boards in the field. In North
America this involves embedding a reinforcing tape, which usually consists of
paper but glass tape is sometimes used as well. The joint cement is normally
based on stucco that is retarded to a certain length of setting time, i.e., up to
90 min. Obviously, these preparations must adhere to the paper, they must not
show shrinkage and they must be able to produce an extremely thin coat to-
wards the edges of the joint. To achieve this, the materials used include latex
binders, thickeners, and non-leveling agents. Examples of latex binders are
polyvinyl acetate and polyethylene co-vinyl acetate. Thickeners may include
ethyl cellulose, hydroxy propyl methylcellulose, methyl hydroxy propyl cellu-
lose, hydroxy ethyl cellulose, and hydroxy propyl cellulose, hydroxy ethyl cellu-
lose and hydroxy ethyl cellulose. Non-leveling agents include attapulgite, sepio-
lite, bentonite, and montmorillonite clays.

7.8 Recycling
In some plants, reject board is ground and used as accelerator. Care must be
exercised for this procedure because of the presence of retarder and because
ground gypsum board is a variable accelerator. Also, reject board is ground and
added to the gypsum feed for the calciner. In recent years, small companies
have emerged that take the rejected board, separate the paper from the core,
and then return the core material to the board plant for use as raw material.
The use of rejected board or otherwise recycled board can be controversial
because it may raise the amount of paper present in the newly formed board to
above 2 %, which is not acceptable for type X board.

References
关7-1兴 Sackett, A., “Inside Wall Covering,” U.S. Patent 520,123 共May 22, 1894兲.
关7-2兴 Wenk, R. J., “Gypsum Board—Opportunities and Challenges,” XIII Eurogypsum
Congress, Salzburg, Austria, May 1987.
关7-3兴 ASTM E119, 2006, “Standard Test Methods for Fire Tests of Building Construction
and Materials,” Annual Book of ASTM Standards, Vol. 04.11, ASTM International,
West Conshohocken, PA.
关7-4兴 Gypsum Association, Fire Resistance Design Manual and Sound Control, GA 600–
2000, Washington, DC.
关7-5兴 Burrows, B., “A Decade’s Experience with Gypsum Board Weight Reduction in the
U.S.,” Proceedings of 14 Internationale Baustofftagung, 1-0197, 20–23 September
2000, Bauhaus University Weimar.
关7-6兴 Q. Yu et al., International Patent Application WO1998US15874 共pending兲.
Gypsum Board 101

关7-7兴 Veeramasuneni, S., Yu, Q., and Frank, F. A., “Gypsum-Containing Composition
Having Advance Resistance to Deformation,” U.S. Patent 6,815,049 B2 共November
9, 2004兲.
关7-8兴 ASTM C1396, 2006, “Standard Specification for Gypsum Board,” Annual Book of
ASTM Standards, Vol. 04.01, ASTM International, West Conshohocken, PA.
关7-9兴 R. Bruce and D. Mitakidis, “Improved Gypsum Board,” U.S. Patent No. 5,085, 929
共1992兲.
关7-10兴 S. Sucech: “Producing Foamed Gypsum Board Using a Foaming Agent Blend,” U.S.
Patent No. 5, 643,510 共1997兲.
关7-11兴 R. A. Kuntze: “Synthetic Plaster Set Retarder,” U.S. Patent No. 3,304,189 共1967兲.
关7-12兴 Jolicoeur, C., “Water Reduction and Reaction Control in Gypsum Processes by Dis-
persing Agents, Short Course,” Gypsum 2000 Conference, May 16–19, 2000, Ortech
International, Toronto.
MNL67-EB/Oct. 2009

8
Glass Mat Gypsum Board
8.1 Sheathing
PAPER LINED GYPSUM BOARD SUCH AS EXTERIOR SHEATHING IS
particularly sensitive to long term exposure to rain and snow in the field. These
problems occur despite the use of a water resistant core and water resistant
front paper. Sheathing is intended to be covered with a weather barrier but it is
often installed and left uncovered. The same situation exists when the board is
stored in the open without adequate protection. For this reason, attempts have
been made to replace the paper with a glass fiber mat.
In North America, Georgia-Pacific 共GP兲 was the first company to develop
and market such a board. The mats are partially covered by a thin layer of core
material pressed through the mat during the manufacturing process, which is
very similar to the production of normal gypsum board. The core can be treated
with hydrophobes, i.e., “water proofing agents,” such as asphalt/wax emulsions,
silicones, or siloxanes usually during the production process. The board meets
ASTM C1177 specifications and it is manufactured as type X board as well
关8-1兴. This type of board is particularly suitable for applications in so-called
exterior insulation and finishing systems 共EIFS兲, where the board is often ap-
plied to the exterior of building and left exposed to rain and snow for some
time. British Plaster Board had earlier developed a similar board made by a
similar process. This board is now manufactured in North America and used
for the same purpose. It differs from the GP board in that the fiber-glass mat is
completely embedded and covered by a thin layer of gypsum. Because of the
successful application of the glass mat board, attempts are underway to replace
it for the paper covered sheathing as an EIFS substrate.
8.2 Backing Board
A similar type of board is used as a water resistant backing board for ceilings
and walls in bath or shower areas. It meets ASTM C1178 and is designed as a
base for the application of ceramic or plastic tile 关8-2兴.
References
关8-1兴 ASTM C1177, 2006, “Standard Specification for Glass Mat Gypsum Substrate for
Use as Sheathing,” Annual Book of ASTM Standards, Vol. 04.01, ASTM Interna-
tional, West Conshohocken, PA.
关8-2兴 ASTM C1178, 2006, “Standard Specification for Coated Glass Mat Water Resistant
Gypsum Backing Panel,” Annual Book of ASTM Standards, Vol. 04.01, ASTM Inter-
national, West Conshohocken, PA.

102
Copyright © 2009 by ASTM International www.astm.org
MNL67-EB/Oct. 2009

9
Gypsum Fiber Board
FOR A VARIETY OF REASONS, ATTEMPTS HAVE BEEN MADE TO PRO-
duce a different type of board by incorporating reinforcement material
throughout the core rather than just applying it to the surface. A number of
these boards have been described in the literature, but only a few have found
application in practice. In general, they are characterized by a considerably
lower rate of production.
For fiber reinforced gypsum board, fiber is usually incorporated through-
out its core to provide improved tensile, flexural, and impact strength. A variety
of fibers, such as paper, wood, glass, or mineral fibers, are normally used for
this purpose. The proportion of paper fiber in this type of board can reach
15 wt. %, which results in a 1 / 2 in. 共12.7 mm兲 board that is relatively heavy at
2100– 3100 lbs MSF 共10.2– 15.1 kg/ m2兲. Although additional surface liners are
normally not required, they are used in some cases. Most efforts to develop a
gypsum board without a paper surface involve the use of paper fibers. Wood
fiber or wafer-reinforced gypsum board is a relative newcomer to the industry,
which was developed because of the need to replace formaldehyde-containing
glue in particleboard. Synthetic polymeric fibers have found little application,
probably because of their high cost.
All gypsum fiberboards are produced in a similar fashion. Differences be-
tween individual processes are related primarily to the methods of mixing,
moistening, and densification the material. Generally the stucco is mixed with
the paper fiber and a relatively small amount of water is added which produces
a “semi-dry” or “wet” compound. This mixture is placed into molds of a very
large sheet press, similar to those used in the manufacture of resin-bonded
particleboard. After the stucco has hydrated, the resulting board is dried using
a method similar to that used for regular gypsum board. The major advantage
of this process is the lower cost for the paper, particularly if waste paper is
used. In addition, drying costs are reduced significantly because the board con-
tains less water, which results of course in higher weights. In regular gypsum
board, drying cost is approximately 20 % while the cost of paper is about 40 %.
The development and application of paper or wood fiberboard occurred
primarily in Europe with only a few operations in North America. There are a
number of companies with very similar technologies. The Ferma Company in
Germany originally developed gypsum fiberboard as Fermacell about 20 years
ago. The technology has improved considerably since that time, but it is still
essentially based on mixing waste paper with stucco. Also, there is always com-
petition from traditional gypsum board companies. For example, Knauf pro-
duces an approximately 4 ⫻ 4 ft2 共1.2⫻ 1.2 m2兲 board for the do-it-yourself
market.
In North America, Würtex and Schenk have provided technology for a plant
constructed by Furman Lumber at East Providence, Rhode Island, and to

103
Copyright © 2009 by ASTM International www.astm.org
104 Gypsum

Louisiana-Pacific for a plant in Nova Scotia that was for a time operated by
U.S. Gypsum. Total construction costs were estimated at U.S. $37 million as
compared to U.S. $28 million for the plant of Louisiana-Pacific in Nova Scotia.
Both plants are no longer in operation.
The Würtex and Schenk plants use less water than Fermacell by employing
a continuous semi-dry process. This permits higher production rates that are
achieved by a continuous operation rather than by pressing of board stacks.
Also, excess water is not removed by the press and hence better dimensional
tolerances are possible and final sanding is not required. Both plants also allow
for the production of a board with a lightweight center of the core by incorpo-
rating expanded perlite. This approach recognizes that normal board from this
process is too heavy for the North American market.
Specifically, the stucco is initially mixed with wetted waste paper supplied
from the U.S. East Coast. Sufficient water is added prior to forming the board
and the remaining water is subsequently spray applied. The plants are designed
for two lines using various mixing and forming heads. The board travels
through this section on a thin belt. Pressing is accomplished with continuous
belt presses whereby one belt must be a sieve to allow for the removal of en-
trapped air. The press time ranges from 30 to 45 min. The wet board then
passed through a jet dryer, which requires about 30 min for drying. The fin-
ished board has a thickness of about 1 / 2 in. 共12.7 mm兲 and weighs slightly
more than 2400 lb per MSF. This is considerably heavier than typical 1 / 2 in.
regular gypsum board, which weighs less than 1700 lb per MSF. The capacity
of manufacturing this type of board is relatively high amounting to approxi-
mately 45 ft/ min based on a board 4 ft 共1.2 m兲 wide.
The manufacturing of Knauf board is based on a process developed by
Hatcheck, which originally used stucco and paper fiber and was later modified
for the manufacture of asbestos cement. For this process, a thin film of stucco
and paper fiber is formed and placed onto a continuous felt where the water is
removed by suction. The film is then wound around a mandrel, which is
stopped when the desired thickness is reached. It is then cut lengthwise to
obtain a sheet and further curing and drying is carried out as appropriate. The
product obtained is relatively dense but can be used for furniture and computer
floor applications.
Gypsum fiberboard products meet gypsum board specifications. They are
covered by a separate standard 关9-1兴. It is also claimed that the finishing of the
joints can be accomplished without the use of reinforcing tape and for this
reason it is faster than for regular gypsum board. However, this is in part a
reflection of the differences in construction practice between Europe and North
America, where reinforcing tape is used to finish gypsum board joints. Addi-
tionally, the surface of the fiber-reinforced board differs since it is composed of
a mixture of fiber and gypsum instead of paper exclusively. Therefore, paints
with especially good hiding characteristics are required for those boards. Gyp-
sum fiberboard has a tendency to be more abuse resistant than regular gypsum
board where the softer core is exposed when the surface paper liner is dam-
aged.
Gypsum Fiber Board 105

References
关9-1兴 ASTM C1278, 2005, “Standard Specification for Fiber-Reinforced Gypsum Panel,”
Annual Book of ASTM Standards, Vol. 149, ASTM International, West Consho-
hocken, PA.
MNL18-67/Oct. 2009

10
Testing, Analysis, and Installation
APPROPRIATE SAMPLING AND TESTING IS AN IMPORTANT CONSID-
eration to obtain materials with consistent qualities throughout the entire
manufacturing process, but particularly during and after calcination. Unfortu-
nately this point is often overlooked or neglected as indicated by the absence of
suitable sampling stations or ports in many locations. This can lead to difficul-
ties particularly when problems are to be solved. Even under normal circum-
stances it amounts to an unsatisfactory situation.
In actual practice, these methods are a blend of procedures developed by a
particular company and procedures published by standards writing organiza-
tions such as ASTM. To some extent, this allows testing of the behavior of the
stucco produced during calcination as well as testing of the stucco slurry dur-
ing board manufacture. For example, certain test methods are intended to
cover situations where the stucco meets the requirements of the specifications
but actually contains a small amount of non-calcined gypsum. The presence of
this effect will not necessarily show up in routine laboratory consistency test
where gentle hand mixing is employed. However, it will show up in actual plant
operation due to the high mechanical action and shearing action of the pin
mixer.
The detection of this behavior can be made in the laboratory by the use of
a high speed mixer such as a “malted milk mixer.” Normal stucco will not
thicken and will pour from the container at the end of the test. However, if the
slurry thickens within 15– 20 s of mixing, the stucco probably contains consid-
erable amounts of gypsum that has not been calcined.
Another method of gauging the initial characteristics of stucco slurry dur-
ing actual plant operation involves measuring the slurry temperature. Tempera-
ture readings are taken directly beneath the pin mixer discharge boot and as
close to the nip of the forming plate as possible.
In actual practice, a temperature rise of 1 ° F 共0.7° C兲 is considered to be
within acceptable limits. If the temperature rise between these two points ex-
ceeds that limit it shows that the setting occurs too early and that paper-to-core
bond problem may result. Hence, the possibility of the presence of uncalcined
gypsum should be determined. This can be done most easily by subjecting the
stucco to DTA or DSC analysis 关10-1兴
To complicate matters, stucco may be calcined beyond the proper level so
that it contains soluble or even insoluble anhydrite, which can cause problems
with set control and paper to core bond. This characteristic will be noticed
during routine quality control tests because its combined water will be below
the specified range. However, if the stucco contains both over-calcined and
under-calcined material, it may give the impression of being acceptable. There-
fore, determinations of the hydration rate of the stucco should be carried out
routinely.

106
Copyright © 2009 by ASTM International www.astm.org
Testing, Analysis, and Installation 107

Routine testing is generally carried out according to methods developed by


standards writing organizations, such as ASTM, British Standards Organiza-
tion, DIN, Canadian Standards Organization, and the Japanese Standards Or-
ganization. These organizations are responsible for establishing a consensus
among a variety of participants prior to publishing the standards. They are
seldom involved in the testing process itself. The standards are similar with the
exceptions that they cover some materials or products that are not manufac-
tured elsewhere. For example, DIN 18163 covers molded wall 共partition兲 panels
that are not mentioned by ASTM because this product is not produced or used
in North America. Also, some locally developed test methods may be preferred,
such as the DIN water wicking method instead of a normal consistency
method.
Standards are usually subdivided into 共1兲 Test Methods, 共2兲 Specifications,
and 共3兲 Application or Installation Procedures. Accessories and related prod-
ucts are treated in the same manner and chemical analyses are normally in-
cluded with test methods. A distinction is usually made between materials and
products although that is not always easy. The test methods are intended to
establish the significant performance characteristics or properties of the mate-
rials and products. These must meet specific criteria that are outlined in appro-
priate specifications usually in a prescriptive rather than a descriptive manner.
Application or installation procedures serve as a best practice guide as to how
the products are to be used in practice.

10.1 Materials Testing


ASTM C11 refers to the terminology related to gypsum and gypsum products
whereas ASTM C22 provides the initial test methods for gypsum to be used to
produce various products. Essentially it establishes that gypsum shall contain
not less than 70.0 wt % of dihydrate, CaSO4 · 2H2O. The chemical analysis of
gypsum and gypsum products is covered by ASTM C471. ASTM C472 outlines
the test methods for physical testing of gypsum, plasters, and gypsum concrete
关10-2兴. It covers methods for stucco, such as free water 共drying兲, fineness 共siev-
ing兲, consistency 共cone penetrometer兲, setting time 共Vicat needle兲, compressive
strength 共2 in., i.e., 5 cm cubes兲, and density 共weighing兲. It is emphasized that
the available material must be tested in a dried condition and that all equip-
ment must be kept free of freshly set plaster. Gypsum veneer plaster is a special
case with ASTM C587 covering joint strength, bond strength, impact strength,
and flexural strength.

10.2 Materials Specifications


The corresponding specifications for plasters are given in ASTM C28 关10-3兴,
which stresses that the physical properties differ in the field when plaster is
applied to lath or masonry and that they are affected by a variety of conditions
experienced in practice. The specification covers neat plaster 共stucco and water
only兲 and plasters containing aggregate such as sand, perlite, vermiculite or
wood fiber. Stucco must contain at least 66.0 wt % of hemihydrate,
CaSO4 · 共1 / 2兲H2O. The specification refers only to the setting time and com-
108 Gypsum

pressive strength of these materials giving the permissible composition 共ratios


of stucco to aggregate兲 where applicable. Fineness is addressed only for gaug-
ing plaster intended for finishing coats. Other examples are specifications for
Casting and Molding Plasters 共C59兲, anhydrous Keene’s Cement 共C61兲, and Gyp-
sum Concrete 共C317兲. The latter covers density instead of fineness.

10.3 Product Testing


For board products, the test procedures are intended to outline sampling and
the measurement of properties as they affect the entire product as produced by
a specific manufacturing process 关10-4兴. These are outlined in ASTM C473.
Fire protection is an example of product testing. It is an undisputed fact
that there is a considerable delay for the back of hardened gypsum board to
reach a failure temperature when it is exposed to an actual fire or a fire test.
One of the hypotheses offered for this protective effect is the formation of a
steam curtain at the front of the board that is created when the dihydrate in the
gypsum dehydrates. This steam curtain allows the temperature to rise only very
slowly. However, this hypothesis is difficult to maintain as the exclusive mecha-
nism in the light of the magnitude of the temperature and length of exposure to
which the board is subjected in a fire test 关Fig. 10.1兴. Other factors must be
taken into consideration.
It should also be noted that the requirements for gypsum board normally
lie within the first hour of the E119 test procedure, that is, failure occurs as the
furnace temperature is still rapidly increasing. Therefore, it is more likely that
the slow progression of the temperature into the gypsum is assisted consider-
ably by the fact that the sponge-like structure of the board prevents crack for-
mation. This assumption is supported by the observation that failure of gypsum
board in a fire test usually occurs when a crack appears in the core through
which hot gases reach the back of the board.

10.4 Product Specifications


In North America the overriding standard is ASTM C1396, which is a recent
combination of several individual standards. Similar test methods are issued by
other standards writing organization 共e.g., DIN 18180兲. It should be understood
that these are minimum requirements and consequently the boards exceed
them in many instances. These tests are again considered routine affairs. The
only contentious issue is the nail pull resistance of the board. For example, if
the values given in C1396, Section 5.1.3 are plotted, a break appears at the
1 / 2 in. 共12.7 mm兲 value. This requires correction since a straight line should
be obtained for gypsum board. This means that the paper contributes to the
nail pull value in a constant manner; say 5 lb 共2.27 kg兲. The core on the other
hand contributes a value which increases with its thickness 共Fig. 10.2兲. It could
be argued that the nail pull values for 5 / 8 and 3 / 4 in. board are too low. How-
ever, this would require exceptionally high nail pull values for both these
boards.
The specifications cover the significant properties of a board such as flex-
ural strength, humidified deflection, core-, edge-, and end-hardness, as well as
Testing, Analysis, and Installation 109

Fig. 10.1—Fire endurance test.

Fig. 10.2—Nail pull test specifications.


110 Gypsum

nail pull resistance. This includes dimensions and tolerances for thickness,
width, ends, and edges of these boards. They also cover criteria which differen-
tiate the type of boards primarily according to the intended use. For example,
the following boards are now covered by ASTM C1396 关10-5兴: gypsum wall-
board and predecorated gypsum board with and without foil backing; gypsum
backing board, coreboard, and shaftliner board; water resistant gypsum back-
ing board; exterior gypsum soffit board; treated core and non-treated core gyp-
sum sheathing board, gypsum base for veneer plasters; gypsum lath, and gyp-
sum ceiling board. Gypsum formboard is a special case covering a permanent
form for poured in place-reinforced gypsum concrete, which also addresses
resistance to fungi. For water-resistant backing board and sheathing the speci-
fication also covers water resistance and surface adsorption. Where foil backed
board is involved, water vapor transmission 共permeance兲 is specified.
In most cases these boards are also produced as type X 共special fire rated兲
board, which must meet the definition for fire resistance of type X board and
type X lath 共E119兲 and flame spread 共E84兲. In practice, these tests are per-
formed and listed by UL or other listing agencies for individual assemblies of
which the board is a key component. This is intended to assure that type X
board meets the requirements of Building Codes. Sound transmission situa-
tions are handled in a similar manner.
This situation gives the impression that there is a good perception of the
fundamental parameters that define the performance characteristics of the gyp-
sum board or one of its components. However, this is not always the case. It is
characteristic for manufacturing to respond to observed shortcomings of the
board by adding yet another additive. This is bound to be problematic in view
of the presence of a large variety of these additives that are used for various
purposes and whose interplay is not known. Little or no attention is being paid
to the fundamental characteristics of the board or its core that make it stiff,
strong, hard, or brittle.

10.5 Testing Problems


Nail pull resistance is a case in point for a remaining problem. As stated pre-
viously, it is the key property that prevents lowering the weight of regular board
below a certain level. Based on fracture mechanics, a hypothesis was advanced
to describe the failure mechanism that occurs when a load is applied to a test
nail in the laboratory inserted into the board during the test leaving the nail
head resting on its surface 共Fig. 10.3兲. Essentially, radial cracking was noticed
first at the bottom of the board at a load of about 30 lb 共13.6 kg兲. Next, at a load
between 30 and 50 lb 共13.6 and 22.7 kg兲 a circumferential crack with a diam-
eter of 1 in. 共2.5 cm兲 was noticed at the back of the board in the core. This
crack then propagates upwards with increasing load from the bottom of the
board towards the edge of the nail head. Further increase in load causes shear
failure to occur immediately under the nail head when a maximum load of
80 lb 共36.3 kg兲 is reached 关10-6兴.
This mechanism is difficult to reconcile with observation during the actual
Testing, Analysis, and Installation 111

Fig. 10.3—Nail pull test assembly.

test on gypsum board and it is also at odds with observation on other brittle
materials. For these materials, a cone is formed as well but it typically propa-
gates from the top, where the nail head impacts the board, towards the bottom.
This behavior has been confirmed for gypsum board by work carried out at the
Research Laboratories of British Plasterboard 关10-7兴. It must be remembered
that hydrated stucco forming the core of gypsum board is brittle and contains
many cracks or fissures, which open under tension but “come up solid” in
compression. As a consequence, gypsum board does not obey Hooke’s law and
it has a correspondingly low Young’s modulus.

10.6 Analyses
The chemical analyses of gypsum and gypsum products are covered by ASTM
C471, which covers gypsum, plaster, and a variety of products ranging from
ready mixed plaster to gypsum board. The individual analytical procedures
address not only the basic constituents such as calcium oxide and sulfur triox-
ide but also a number of impurities such as chlorides. Analyses of FGD gypsum
are of course somewhat more involved. These points are discussed in the pub-
lication entitled “Analysis and Evaluation of Synthetic Gypsum” 关10-8兴. A Ger-
man publication discusses the Analysis of FGD Gypsum in a more detailed
manner 关10-9兴.

10.7 Application and Installation Procedures


A variety of standards have been issued that are intended to cover the applica-
tion of these materials, boards, and accessories in the field 关10-10兴. Testing,
specification, and application/installation standards should therefore be con-
sidered an integral part of the entire process of production and construction.
This assumes appropriate feedback from the field. While some testing accord-
ing to specifications is carried out in the manufacturing plant, other tests are
carried out by approved agencies and are performed according to installation
standards.
On the surface this may appear to be a static procedure but in reality it
involves new developments and changes on an almost perpetual basis. An in-
teresting example of this is the fact that the paper liner for sheathing has been
112 Gypsum

replaced by a glass mat, a significant change in the structure of this board. Also,
the paper liner of the gypsum board can be wetted and the board can then be
applied to cover rounded surfaces such as pillars and arches. In these cases the
original strength is largely restored by drying 关10-11兴.

References
关10-1兴 Kuntze, R. A., “The Determination of Small Amounts of Gypsum in Calcium Sul-
fate Hemihydrate by DTA,” Materials Research and Standards, ASTM International,
West Conshohocken, PA, 1962, p. 640.
关10-2兴 ASTM C472, 2006, Annual Book of ASTM Standards, Vol. 04.01, ASTM Interna-
tional, West Conshohocken, PA.
关10-3兴 ASTM C28, 2006, Annual Book of ASTM Standards, Vol. 04.01, ASTM International,
West Conshohocken, PA.
关10-4兴 ASTM C473, 2006, Annual Book of ASTM Standards, Vol. 04.01, ASTM Interna-
tional, West Conshohocken, PA.
关10-5兴 ASTM C1396, 2006, Annual Book of ASTM Standards, Vol. 04.01, ASTM Interna-
tional, West Conshohocken, PA.
关10-6兴 Peterson, K., “Engineered Gypsum Panels,” Sixth International Conference on Gyp-
sum, Ortech, International, Toronto, May 2000.
关10-7兴 Private communication from BPB Research Centre, London.
关10-8兴 Luckevitch, L. and Kuntze, R. A., Sixth International Conference on Gypsum,
Ortech, International, Toronto, May 2000.
关10-9兴 Analyse von REA-Gips, VGB-Merkblatt, 1991.
关10-10兴 ASTM C840, 2006, Annual Book of ASTM Standards, Vol. 04.01, ASTM Interna-
tional, West Conshohocken, PA.
关10-11兴 US Gypsum Company/CGC Inc., Construction Handbook 1993.
MNL67-EB/Oct. 2009

11
Health Considerations
GYPSUM IS CONSIDERED TO BE A NON-TOXIC MATERIAL WITH THE
dust falling into the nuisance category. Natural gypsum 共not synthetic gypsum兲
is recognized as calcium sulfate to be acceptable for human consumption and a
wide variety of other uses by the U.S. Food and Drug Administration. It is a
dietary source for calcium, it is used as color additive for drugs and cosmetic, it
conditions water used for beer brewing, it controls the tartness and clarity of
wine, and it is an ingredient in canned vegetables, flour, white bread, ice cream,
blue cheese, etc.
As indicated by the term landplaster, gypsum is used as a fertilizer since it
benefits a variety of crops where substantial amounts of sulfate sulfur are re-
quired. It is also used as soil additive to improve the workability, water penetra-
tion of soils, and to overcome the effect of alkalinity. It settles dirt and clay
particles in turbid water, particularly ponds, without injury to aquatic life. Fi-
nally, surgical and orthopedic casts are often made from hydrated ␣-plaster, i.e.,
gypsum 关11-1兴.
NIOSH requires a maximum REL of 10 mg/ m3 TWA and 5 mg/ m3 共respi-
ratory兲, whereas OSHA allows a maximum PEL of 15 and 5 mg/ m3, using
NIOSH methods 0500 and 0600. NIOSH also specifies first aid procedures in
case of contamination of eyes, respiratory system, or by ingestion 关11-2兴.
Of course, impurities of the gypsum have to be considered separately and
individually. For example quartz 共determined by DTA兲 must not be present in
quantities that exceed EPA requirements.

References
关11-1兴 Gypsum’s Environmental Story, Gypsum Association, Washington, DC.
关11-2兴 NIOSH Pocket Guide to Chemical Hazards, 2005, Washington, D.C.

113
Copyright © 2009 by ASTM International www.astm.org
MNL67-EB/Oct. 2009
Bibliography
The following literature has also been used for the preparation of this manuscript:

关1兴 Proceedings of the New Frontiers for Byproduct Gypsum, Ortech Corporation, Tor-
onto, 1988, Ortech International.
关2兴 Proceedings of the Second International Conference on FGD and Chemical Gypsum,
Ortech Corporation, Toronto, 1991, Ortech International.
关3兴 Proceedings of the Third International Conference on FGD and Chemical Gypsum,
Ortech Corporation, Toronto, 1993, Ortech Corporation.
关4兴 Proceedings of the Fifth International Conference on FGD and Synthetic Gypsum,
Ortech Corporation, Toronto, 1997, Ortech Corporation.
关5兴 Gordon, J. E., The New Science of Strong Materials, Penguin Book A920, 1968, Balti-
more.
关6兴 GA-600 Fire Resistance Design Manual and Sound Control of Gypsum Systems,
Gypsum Association, Washington, DC.
关7兴 Desch, C. M., Henri Le Chatelier 1850-1936, Memorial Lecture/The Chemical Society
1933–1942, London, Vol. IV, 1951.
关8兴 Wirsching, F., Calcium Sulfate, Ullmann’s Encyclopedia of Industrial Chemistry, 5th
ed., 1985, VCH Verlag, Weinheim, Germany.
关9兴 Oetzel, M., “Studies of the Phase-Connections in the System CaSO4 – H2O,” Ph.D.
dissertation, RWTH Aachen, Germany.
关10兴 Bezou, C., Nonat, A., Mutin, J.-C., Christensen, A. N., and Lehmann, M. S., “Investi-
gation of the Crystal Structure of ␥-CaSO4, CaSO4 · 1 / 2H2O, and CaSO4 · 0.6H2O
by Powder Diffraction Methods,” J. Solid State Chem. Vol. 117, 1995, pp. 165–176.
关11兴 Shchukin, E. D. and Bright, J. E., , “Bridging of Crystals in the Process of Hydration
Hardening of Gypsum. Hydration and Setting of Cements,” RILEM, E&FN Spon,
London, 1992.
关12兴 Kuntze, R. A., “The Chemistry and Technology of Gypsum 共Symposium Proceed-
ings兲,” ASTM Special Technical Publication, 1984.
关13兴 Gips-Datenbuch, Bundesverband der Gips- und Gipsbauplattenindustrie e. V.,
Darmstadt, 1995 共in German兲.
关14兴 Weist, J. A., Kuntze, R. A., Luckevich, L. M., Collins, R. E., and Theodore, F., “The
Gypsum Industry and Flue Gas Desulfurization 共FGD兲 Gypsum Utilization, A Util-
ity Guide,” Report No. EPRI TR-103652, 1994.
关15兴 Murat, M., Structure, Christallochemie and Reactivity of Calcium Sulfates, RILEM
International Symposium, Saint-Rémy-Les-Chevreuse, 1997.

114
Copyright © 2009 by ASTM International www.astm.org
www.astm.org
ISBN #: 978-0-8031-7015-5
ASTM Stock #: MNL67

You might also like