Engineering investigation including insitu positive and negative lateral load testing (out of plane) of composite wall panels (clay brick and concrete block masonry).
In the Spring 1979, a long length of full height exterior masonry wall panel from the second storey of a two storey building at Scott Maritimes Ltd., Abercrombie Point, Nova Scotia collapsed onto and through the roof of an adjacent one storey building of the Scott Maritimes Ltd. (Scott) wood pulp manufacturing plant. At the time of the wall collapse, the wall was subjected to insignificant wind suction force.
For the most part, the exterior walls of all buildings at the Scott plant consist of non-reinforced 12 in. (nominal) thick masonry walls constructed of 8 in. (nominal) thick hollow concrete block masonry units bonded with 4 in. (nominal) thick clay face brick (see photograph no. 7).
The lateral support for the masonry walls was intended to be provided by the installation of hot dipped galvanized mild steel strap anchors having one end embedded into the masonry wall construction while the other end to prevent outward movement was bent at 90° to bear on the flanges of the adjacent interior structural steel columns.
In 1979, J.W. Cowie Engineering Ltd. (Cowie) was retained by Scott to; (1). Establish the cause of the wall collapse, (2). Establish the strength and stability of the remaining masonry walls of the Scott complex, (3). Where required, provide temporary lateral restraint details for immediate installation, and (4). Provide permanent restoration design details for tendering.
As a consequence of Cowie's initial investigation, it was discovered the mild steel hot dipped galvanized lateral support strap anchors were improperly installed as revealed by the following defects;
(a). A gap existed between the bearing ends of the anchor at the face of the steel flanges of the columns thereby providing no lateral restraint against the outward movement of the walls when subjected to wind suction forces.
(b). A gap existed between the interior face of the concrete block masonry wall and the steel column flanges hence, providing no lateral restraint against inward movement when subjected to wind pressure. In some instances, both defects of the lateral restraint anchorages existed at the same column hence, providing no lateral restraint for either positive or negative wind force.
In addition to improper installation of the strap anchors, the strap anchors were failing as a consequence of brittle fracture at the 90° bend for bearing on the column flange. Many of the bearing ends of the strap anchors could be easily removed from the column flanges by minimal finger force. Brittle fracture
failure was most likely the result of deterioration due to exposure to the corrosive environment within the wood pulp manufacturing facility.
In addition to the lateral restraint anchorage problems, portions of the exterior masonry walls had deteriorated due to moisture migration from the interior side of the wall to the exterior of the wall and subsequent freeze/thaw action.
Throughout a number of wall areas, the clay brick headers failed in shear as a consequence of differential movement between the brick masonry and the concrete block masonry. Non recoverable moisture expansion of the brickwork and the drying/shrinkage strains within the concrete block masonry caused sufficient shear force to crack the header bricks anchoring the brickwork to the blockwork.
As a result of the shear failure of the header bricks, the composite action between the clay brick masonry and the concrete block masonry was lost. In wall areas where the composite action was lost the masonry walls were demolished and reconstructed.
In general, the lateral restraint of the exterior masonry walls throughout all buildings of the Scott plant were considered defective and required the installation of new Cowie custom designed stainless steel through wall lateral support anchors.
In addition to lateral restraint defects, the flexural capacity of the walls to resist NBCC design wind loading was in question because of wall openings and excessive span to thickness relationships beyond the traditional prescriptive design criteria.
To achieve the required bending resistance to resist the minimum required NBCC design wind loadings, would necessitate taking into account two way bending action (vertical spanning and horizontal spanning), as well as continuity at lateral support locations.
To make the restoration design more complicated, the installation of new lateral restraint anchors on the interior side of the walls was restricted at many locations by mechanical equipment etc. mounted on or located adjacent to the structural steel columns and spandrel beams.
At the time of Cowie's investigation, there were no NBCC/CSA recommended analysis and design criteria for masonry walls subjected to uniformly distributed lateral loading with the walls simply supported on more than two bearing edges.
Cowie chose to follow a method of analysis and design (modified yield line criteria) established by the British Standard Institution (BSI) Standard BS5628 "Code of Practice for Structural Use of Masonry" (Part I "Unreinforced Masonry") published in 1978.
Before the BSI Standard BS5628 analysis and design procedures could be implemented, it was necessary for Cowie to establish the characteristic ultimate bending moments of the composite masonry walls at Scott Maritimes.
Insitu ("out of plane") lateral loading was required in order to establish the ultimate vertical spanning bending moments (normal to the bed joints) and the ultimate horizontal spanning bending moments (parallel to the bed joints). Ultimate bending moment failure in the brickwork (tensile failure) differs from the ultimate bending moment failure in the blockwork (tensile failure). It there- fore was necessary to subject the test walls to both positive and negative lateral loading for both vertical spanning and horizontal spanning test panels.
Cowie retained the testing services of the Heavy Clay Division of the British Ceramic Research Association (BCRA), Stoke-on-Trent, England, to conduct the insitu lateral load wall testing at the Scott plant.
Cowie designed the structural steel support system and reaction frames for the insitu testing of the test walls shown in photographs; 1 to 6. BCRA senior engineer, H. Roy Hodgkinson and his assistant, W. Webb accompanied Cowie to the Scott plant and conducted the lateral load testing. The air bags, pressure gauges and electronic equipment for measuring and recording wall deflections were provided by BCRA. Scott provided compressed air and labour as required.
A total of ten insitu lateral load ("out of plane") tests were conducted consisting of the following:
(a). Four two way spanning tests (two pressure, two suction) of panel size 7 ft. x 6 in. by 7 ft. x 6 in.
(b). Four vertical spanning tests (two pressure, two suction) of panel size 4 ft. 0 in. x 4 ft. 0 in.
(c). Two horizontal spanning tests (one pressure, one suction) of panel size 5 ft. 6 in. x 5 ft. 6 in.
As a result of the insitu lateral load testing, the characteristic ultimate bending moments for both vertical and horizontal directions were established for the 12 in. (nominal) thick composite masonry walls at the Scott plant.
Knowing the ultimate characteristic vertical and horizontal spanning bending moments and with the use of the BS5628 moment coefficients for the various wall heights to length relationships with various edge support conditions, the
ultimate resisting bending moments for the masonry panels having two sided, three sided and four sided edge supports could be calculated. The ultimate bending moments were reduced by 50% for working stress design.
Cowie retained the consulting engineering services of Mr. Barry A. Haseltine, FIStructE. and Mr. Neil J. Tutt, MIStructE. of Jenkins and Potter Consulting Engineers, London, England for assistance in implementing BSI Standard BS5628 used in the development of design curves to take into account wind pressure, panel size and the bearing edge conditions for the various wall panels at the Scott plant. Both Messrs. Haseltine and Tutt played a significant part in the development of BSI Standard BS5628 related to the analysis and design of masonry walls subjected to uniform lateral loading.
Cowie further developed the BSI Standard BS5628 design criteria to enable Cowie to provide restoration design drawings for hundreds of individual masonry wall panels, showing the locations for the installation of the Cowie custom design stainless steel lateral support anchors for two, three and four sided bearing edges. For the most part, the required moment of resistance was achieved by allowing the panel to be supported on more than two bearing edges. There were however, instances where additional steel framing had to be installed to provide lateral support in the areas where the interior surface of the wall adjacent to steel columns and spandrel floor beams was obstructed by mechanical equipment etc.
Cowie also retained the consulting engineering services of C.T. (Tom) Grimm, PE of the University of Texas at Austin, Texas, USA. Cowie removed brick masonry prisms from various wall areas of the Scott plant for testing by C.T. Grimm and to provide advice related to clay brick masonry moisture expansion problems.
It is interesting to note, CSA Standard S304.1-04 "Design of Masonry Structures" recommend a procedure for the analysis and design of uniformly loaded nonreinforced masonry walls having support on more than two edges. Section 7.12 "Flexural Wall Panels" provides a similar analysis and design criteria for wall panels having more than two edge support as that recommended in 1978 by British Standards Institution Standard BS5628, Part I, "Code of Practice for Structural Use of Masonry, Part I Unreinforced Masonry". CSA Standard S304.1-04 however, has only three wall panels with varying edge bearing conditions whereas, BSI Standard BS5628 has twelve panels with varying edge bearing conditions and associated bending moment coefficients.
01. Structural steel portal frames installed for support of the masonry wall above the horizontal opening within the wall. The portal frames were installed prior to removing the horizontal section of masonry wall. The masonry wall from the horizontal opening to foundation level was subsequently saw cut into rectangular panels for lateral 'out of plane' load testing to failure. A typical reaction frame is shown placed against the exterior face of the brickwork. The reaction frame is to provide bearing edge support on the brickwork of the masonry wall when the wall is subjected to outward test pressure against the interior face of the wall.
02. Structural steel framing installed for support of the masonry wall above the wall area to be saw cut into panels for lateral load testing. Also shows the installed Cowie custom designed stainless steel anchors for anchorage of the masonry wall to the adjacent structural steel framing of the building. Shows the masonry wall being saw cut into panels for lateral load testing.
03. Reaction frame (steel framing and plywood sheathing) placed adjacent to the exterior face of the brickwork of the saw cut masonry wall panel. Air bags were placed between the plywood sheathing of the reaction frame and the face of the brickwork. Inflation of the air bags at increasing pressure provided a uniform pressure on the brickwork synonymous with 'out of plane' wind loading.
04. Shows the interior face of the concrete block masonry of the solid wall construction. The face of the concrete block masonry has been painted with 'white wash' to enhance the documentation of cracking within the blockwork caused by the inward pressure on the face of the brickwork. Also shows the location of the linear transducers used to measure the inward deflection of the blockwork.
05. Shows the reaction frame mounted adjacent to the face of the concrete blockwork of a test wall panel. Air bags were placed between the plywood sheathing of the reaction frame and the face of the blockwork. Inflation of the air bags represented a uniform outward pressure synonymous with wind suction force.
06. Shows the installation of linear transducers for the measurement of the deflection of the brickwork as a consequence of the outward force of the air bags on the back face of the test panel.
07. Typical section through wall 12 in. (nominal) solid, composite exterior masonry wall construction. The wall construction consists of 4 in. (nominal) wide clay facing brick with 8 in. (nominal) wide hollow interior blockwork. The brickwork and the blockwork are tied together by brick header courses. Wire ladder type drying/shrinkage reinforcing is installed in every horizontal joint in the block- work. It should be noted, the interior wire of the ladder wire bed joint reinforcement is not placed in the bed joints between the face shells of the blockwork but rather straddles the hollow core openings within the blockwork. The improperly sized and misplaced ladder wire reinforcement contributed little to controlling the drying/shrinkage of the blockwork.