Materials and Structures/Matdriaux et Constructions, 1989, 22,280-286 Evaluation of concrete spall repairs by pullout test F. G. COLLINS Taywood Engineering Ltd, 7/F, 275 Alfred Street North, North Sydney, N.S. W. 2060, Australia H. ROPER School of Civil and Mining Engineering, University of Sydney, 2006 N.S. W., Australia Spalling concrete was simulated in the laboratory by utilizing pullout test methods generally used for the determination of in situ concrete strength. The cracking patterns displayed by pullout test specimens typify the damage of concrete by spalling. All specimens were damaged by pullout testing, repaired with epoxy mortar and subjected to a second pullout test at a later time. The test programme showed that the overriding factor which governs successful repairs to concrete is the soundness of the repair plane. 1. INTRODUCTION A spall is characterized by the breaking away of a fragment, detached from a larger concrete mass by a blow, the action of weather, or by pressure which may result from expansion of porous aggregates, occluded organic material, reactive contaminants, or corroding steel reinforcing bars and/or embedments. For this study, spalling concrete was simulated in the laboratory by utilizing pullout test methods generally used for the determination of in situ concrete strength. The radial and circumferential cracking patterns dis- played by pullout test specimens typify the damage of concrete by spalling. The pullout cavity shape is repro- ducible using constant apparatus geometry. 2. TEST METHOD In order to conduct a pullout test a metal insert is embedded in fresh concrete. After the concrete has hardened, the pullout strength is determined by measur- ing the maximum force required from a loading device to pull the insert and adjoining concrete from the concrete insideDiameter of Reaction Ring__ I l~ Reaction Ring V7 Idealized " ~ Failure [ Surface I Reaction Ring ~ ~Embedment V--,-----_.A pex _I Ang(e 2o~ [ Disk Diameter Fig. 1 Schematicrepresentation of pullout test. 0025-5432/89 9 RI LEM mass. The configuration of the embedded insert and bearing ring of the loading system determine the approximate shape of the pullout fracture within the concrete mass. A typical configuration is shown in Fig. 1. 3. THE PULLOUT TEST AS A MEANS OF EVALUATING CONCRETE STRENGTH The concept of a pullout test as a means of determining the concrete strength in a completed structure was sug- gested by Skramtajew [1] in 1938. Kierkegaard-Hansen [2] conducted a further series of tests to investigate the effect which a bearing ring of finite radius would have on the ultimate pullout force. These tests indicated that for smaller reaction rings (and thus smaller apex angles, 2a) the pullout force increased. Since less material was involved in the failure surface, Kierkegaard-Hansen concluded that the increase in ultimate load was due to a change in the internal state of stress within the failure region. Subsequent studies were aimed at developing calibration curves for various concretes using a constant test geometry. In the early 1970s a series of tests were conducted by Malhotra [3], Richards [4], and Rutenbeck [5] to estab- lish correlations between the pullout strength of con- crete and strength parameters derived from other non- destructive test methods. These tests form the basis of the current ASTM requirements for the pullout test, ASTM C900 [6]. The ASTM requirements define the apex angle to fall between 54 and 70~ A linear correla- tion was found to exist between pullout strength of concrete and compressive strength of concrete cylin- ders. Stone and Giza [7] sought to obtain a lower coefficient of variation for the pullout test by examining different geometries, aggregate types and aggregate sizes. Failure surfaces were found to exhibit a conic frustum geometry at low apex angles and a trumpet-shaped geometry at higher apex angles.