624.152 0r6g EiiNCINSERtNG LtBKARY WWflVEk'SIVY OF SLUWOHS lUJfMMS 6ie0i O’ROURKE Q the ground movements related ■ TO BRACED EXCAVATION AND THEIR INFLUENCE ON ADJACENT BUILDINGS DLC 2? ZOflt DEC 0 7 2004 JAN 1 9 2006 Digitized by the Internet Archive in 2020 with funding from University of Illinois Urbana-Champaign https://archive.org/details/groundmovementsrOOorou PB-267 311 THE GROUND MOVEMENTS RELATED TO BRACED EXCAVATION AND THEIR INFLUENCE ON ADJACENT BUILDINGS Illinois Univ at Urbana-Champaign Dept of Civil Engineering Prepared for: Department of Transportation, Washington, D. C, Office of the Se eretary August 1976 DISTRIBUTED BY: National Technical Information Service U. S. DEPARTMENT OF COMMERCE 5285 Port Royal Road, Springfield Va. 22151 This document has been approved for public release and sale. PB 267 311 Rdport No. DOT-TST 76T-23 THE GROUND MOVEMENTS RELATED TO BRACED EXCAVATION AND THEIR INFLUENCE ON ADJACENT BUILDINGS AUGUST 1976 FINAL REPORT Prepared for U.S. Department of Transportation OFFICE OF THE SECRETARY AND FEDERAL RAILROAD ADMINISTRATION Washington, D.C. 20590 REPRODUCED BY NATIONAL TECHNICAL INFORMATION SERVICE U.S. DEPARTMENT OF COMMERCE SPRINGFIELD, VA. 2Z16I C i Tachnical Report Documontotion Pago 1. Report No. DOT-TST 76T-23 2- Government Acce»»ion No. The Ground Movements Related to Braced Excavation and Their Influence on Adjacent Buildings 3. Recipient * Cotolog No. 5. Report Dote August 1976 6. Performing Orgoniiotion Cod* T. D. O'Rourke, E. J. Cording, M. Boscardin 9. Performing Orgonixotion Nome ond Addre** ^ Department of Civil Engineering University of Illinois at Urbana-Champaign Urbana, Illinois 61801 12 Sponsoring Agency Nome ond Addre,, U.S. Department of Transportation Office of the Secretary, and Federal Railroad Administration Washington, D. C. 20590 _ 15 Sopp I •mentory Note* 8. Performing Orgoniiotion Report No. UILU-ENG-76-2023 10 Work Unit No. (TFiAIS) 11. Confroct or Grant No. DOT FR 30022 13. Type of Report and Period Covored Final August 75 - August 76 14. Sponioring Agency Code 16 . Abitrocf report sunnarizes the settlement and lateral displacement measurements associated with urban excavation projects in the dense sands and interbedded stiff clay of Washington, 0. C. and the soft clay of Chicago. The ground movements caused by excavation in each area are discussed in light of the soil profile and construc¬ tion techniques. The relationship between soil displacement and the damage caused to adjacent buildings is examined. Criteria for the onset of architectural damage are recommended for brick-bearing wall and frame structures subject to excavation movements. Brick-bearing wall structures are described, with special emphasis on the construction details related to building stability. Various modes of instability caused by differential ground movements are examined for brick-bearing wall struc¬ tures. Case histories of building damage caused by adjacent excavation are presented 17. K.y Word, Excavation, Cut-and-Cover Construction, Soil Displacements, Bracing, Building Damage, Underpinning IB. Oi»Y»ibutio« , Document is available to the public through the National Technical Information Service, Springfield, Virginia 22151 19. Sooutify Cla,,il. (ol tbi, report) Unclassified 20. Soewrity Cla,,if. (of thi, po 9 ») Unclassified 22. Prie* po Ae)71 A4.0 >19.2) 60-80 ,20 ’(6.1) 40 ( 12 . 2 ) .60 ’( 18 . 3 ) _80 ( 24 . 4 ) ,100 ’( 30 . 5 ) ,120 ’( 36 . 6 ) Symbols i N - Standard Penetration Resistance, Blovrs/ft Cu-Undrained Shear Strength, KSF (fj, x 10* ) Fig. 2.1 Soil Profile for Braced Excavations in Washington, D. C. Depth,ft (m.) 2-5 Pleistocene deposits was found at a depth of approximately 40 ft (12.2 m) below the street surface. During construction the granular soils of the Pleistocene and Cretaceous strata were dewatered below the subgrade of all braced cuts under study. 2.2.2 SURFACE SETTLEMENT AND CONSTRUCTION METHOD The surface settlements associated with 6 different sections of braced excavation are summarized according to the methods of construction and plotted in dimensionless form in Fig. 2.2. The settlement and distances are expressed as percentages of the maximum excavation depth. Each excavation is listed according to its location and prominent characteristics in a table that accompanies the data plot. Previous research (34) has shown that ground movements generated by strut removal during subsurface construction can result in a 100 percent in¬ crease over the displacement caused by excavation to subgrade. Consequently, the total excavation history, including strut removal and backfilling, should be used as a baseline from which to judge the influence of soil movements on nearby structures. For this reason, the settlement data assembled in Fig. 2.2 represent both the initial and final stages of construction. The construction procedure for each excavation was similar. The excavations were supported by cross-lot braces with vertical separations that averaged between 12 and 16 ft (3.7 and 4.9 m). The braced cuts were deepened approximately 20 to 25 ft (6.1 to 7.6 m) below the lowest, previously installed struts before installation of the next brace level. The excavation walls were \ 2-6 Distance FromExcov. Max. Depth Of Exco^ Case Symbol Location Max. Deoth . ft Support 1 • 7 th a G St., N.W. 60 Soldier Pile-Logging With Cross-Lot Struts For All Coses 2 0 9th a G St., N.W. 60 3 A nth a G St., N.W. 55 4 A 12 th a I St, N.W. 60 5 ■ 7 th a G St., N.W. 80 6 0 1 st a 0 St., s. E. 80 Fig. 2.2 Suitmary of Settlements Adjacent to Braced Cuts in Washington, D. C. 2-7 composed of soldier piles, ranging in size from 14BP102 to 24WF120, on roughly 7 ft (2.1 tn) centers with oak lagging installed between adjacent piles. Prob¬ lems with running or sloughing soil were not encountered for the data indicated. Occasionally, local difficulty in controlling the ground water was experienced but the majority of opencutting was performed without substantial seepage and in soil that locally maintained its vertical face when exposed during lagging operations. When compared with the settlements summarized by Peck (36) excava¬ tions in sand and soft to hard clay, the settlements shown in Fig. 2.2. are small. Expressed as a percentage of the maximum braced cut depth, they range from a value of 0.3 percent near the edge of excavation to values less than 0.1 percent at distances from the edge of excavation equal to or exceeding the maximum depth of the cut. Settlements reported in the literature for excavation in San Francisco (2,42), Los Angeles (28), Boston (27), and Minneapolis (25) that were extended through profiles of medium to dense sand and sand with inter- bedded stiff clay also fall within the same zone of settlement. Figure 2.3 shows the angular distortion associated with braced cuts in Washington, D. C. plotted as a function of the dimensionless distance from the edge of excavation. The angular distortion was estimated by dividing the differential settlement of two points along a line perpendicular to the edge of excavation by the distance separating them. The dimensionless distance associated with each value of angular distortion was estimated as the distance from the edge of excavation to a point midway between the two points from which the differential settlement was computed. The values of angular distortion shown in Fig. 2.3 were derived from the settlement measurements associated i Angular Distortion, 8v x 10' 2-8 Distance From Edge Of Cut Depth Of Cut Case Symbol Max. Depth, ft (m) Support 1 • 60 (10.3) Soildier Pile- Logging With Cross-Lot Struts For All Coses 2 0 60 (18.3) 3 A 55 (16.8) 4 A 60 (18.3) Fig. 2.3 Angular Distortion for Braced Cuts in Washington, D. C. 2-9 with the excavations listed as Case 1 through 4 in Fig. 2.2 The angular dis¬ tortion ranges from a maximum value of approximately 5 x 10’^ near the edge of excavation to a value slightly larger than 1 x lO'^ at a distance from the edge of excavation equal to the maximum depth of the cut. 2.2.3 LATERAL GROUND MOVEMENTS Cut-and-cover subway construction can be visualized as occurring in three prominent stages. During each stage, construction methods and support application contribute to the ground displacement in a characteristic way. These stages include: 1. Excavation before installation of braces. 2. Excavation to subgrade after upper braces installed. 3. Removal of braces and construction of the permanent structure. Figures 2.4 through 2.6 trace the development of lateral soil move¬ ment as a function of the construction history for a cut-and-cover excavation in Washington. D. C. The excavation was 60 ft (18.3 m) deep and was supported by soldier pile and lagging walls with 5 levels of cross-lot struts. A de¬ tailed study of this excavation and others in Washington. D. C. has been per¬ formed by O'Rourke (34). The figures show the horizontal wall movements and lateral distortion in the retained soil for each of the construction stages listed above. The lateral distortion has been estimated from inclinometer measurements by dividing the differential lateral displacement between two points at a given elevation by the distance separating them and plotting the lateral distortion at the mid-point between the two measurements. The contour interval for lateral distortion has been chosen as 0.5 x 10"^. which is the 2-10 critical tensile strain for concrete and masonry elements (38). The lateral strains estimated for this particular excavation compare favorably with inclinometer data from other excavations in Washington, 0. C. The lateral movements and construction sequence are discussed as follows: befor e installation of braces: Because time was required for the connecting of steel lacing between adjacent street beams, the excavation was advanced to a depth of 20 to 25 ft (6.1 to 7.6 m) before the street beams were shimmed against the walls of the cut. Hence, defomation of the wall occurred primarily as a cantilever-type movement. The lateral dis¬ tortion indicated in Fig. 2.4 reflect this mode of defomation showing tensile strains that develop from the upper edge of the cut in a triangular pattern of contours that decrease in magnitude with depth and distance from the wall, soil movements were measured at a distance from: the edge of cut approximately equal to the depth of measured displacement at the excavation wall. This depth exceeds the depth of the excavation bottom by approximately 10 ft (3 m). 2 . rvravatinn to s.iborade aft er upper braces instal.l _ e j.: As the upper braces were installed, the upper portion of the excavation wall was restrained from further lateral movement. In fact, preloading of the upper level struts resulted in a net decrease in the measured lateral movement near the top of the cut. The incremental distortion, plotted in Fig. 2.5, show that there was a recompression of the soil near the top of the excavation as referenced to the soil movement during stage 1 construction. In the deeper portions of the cut, the inward bulging of the excavation wall was associated with lateral tensile strains. These strains, shown in Fig. 2.5, emanate from a section of the wall 2 -n Lattral Displacemtnt, in. (cm) ( 3 . 0 ) ( 2 . 0 ) ( 1 . 0 ) 1.2 0.8 0.4 0 I I In 0 20 ( 6 . 1 ) 40 ( 12 . 2 ) 60 ( 18 . 3 ) 80 ( 24 . 4 ) a. & Fig 2.4 Lateral Distortion Corresponding to Stage 1: Excavation Before Installation of Braces Lateral Displocement, in. (cm) ( 30 ) ( 2 . 0 ) ( 1 . 0 ) 1.2 0.8 0.4 0 * I I I ^ 20 16 . 1 ) E 40 ( 12 . 2 ) o 60 ( 133 ) 80 ( 24 , 4 ) \ N, Fig. 2.5 Lateral Distortion Corresponding to Stage 2: Excavation to Subgrade 2-12 Lateral Displaoetnent, in. (cm) (yo) (2.0) (1.0) 1.2 0.8 0.4 0 * I 20 ( 6 . 1 ) 40 ( 12 . 2 ) 60 ( 18 . 3 ) 80 ( 244 ) Fig. 2.6 Lateral Distortion Corresponding to Stage 3: Removal of Braces Depth,ft (m) 2-13 near the bottom of excavation in contours that loop upward from their point of origin toward the ground surface at an angle of roughly 45° from the verti¬ cal. Near the excavation wall, lateral ground strains were measured at a depth of 1.20 times the maximum depth of excavation. 3. Removal of braces and construction of the permanent structure : As the bottom braces were removed to build the underground structure, further inward bulging of the wall occurred at the lower levels of excavation. When the upper two brace levels were removed, the wall was supported in its lower portion by the subway structure and the corresponding movements resulted from a cantilever-type deformation of the wall. Consequently, the incremental strains are a composite of the soil distortions associated with inward bulging at depth and cantilever movement near the top of the wall. This is indicated in Fig. 2.6 where the plot of the incremental distortions, as referenced to the stage 2 construction, show incremental tensile strains concentrated near the bottom of the cut in a prominently curved pattern of contours as opposed to a nearly triangular pattern of contours at the upper portion of the cut. The increase in ground distortion at levels below the subway invert is very small. The cumulative distortion shown in Fig. 2.6, is the sum contribution of strains developed in the three stages of construction. Two zones of dis¬ tortion can be distinguished. Lateral tensile strain of over 2.5 x lO'^ is concentrated near the wall at the invert line of the subway structure. From this point lateral strains emanate in a looped pattern of contours that are directed upward toward the ground surface. This zone forms the lower boundary 2-14 of ground movement for the retained soil mass. Near the top of the cut the contours of lateral strain are inclined diagonally to the excavation wall and reflect the cantilever movements that have been concentrated at this level of the excavation. The lateral distortion at the ground surface is approximately 2 X 10"^ within a distance of 25 ft (7.6 m) from the edge of excavation. A notable point of deformation occurs at a depth of 30 ft (9.2 m) where the cumulative lateral distortion is zero. This apparent absence of strain re¬ flects the characteristic S-shape of the wall displacements. The conspicuous indentation of the wall at this level derives from the high preloads (approxi¬ mately 60 to 70 percent of the design load) that were jacked into the struts at this level. The struts were preloaded against the clay stratum and the resulting plastic deformation was not fully recovered when the struts were r6inov6cl duririQ stsQ® ^ construction. In sunmary, the lateral movements generated by opencutting are re- lated directly to the mode of deformation at the excavation wall which, in turn, is related to the construction procedure. Three stages of construction (initial excavation before placement of braces, excavation to subgrade, and removal of struts) can be distinguished for cut-and-cover excavation that contribute to the ground displacements. The final displacement profile is a composite of the soil movements generated during each stage of construction and is composed essentially of two zones of lateral distortion. A deep-seated zone of lateral strain develops in response to inward bulging of the excava¬ tion wall and forms the lower boundary of the mobilized soil mass. An upper zone of lateral distortion develops in response to cantilever movement of the excavation wall and contributes directly to the surface strains within a dis¬ tance of 20 to 30 ft (6.1 to 9.2 m) from the edge of the cut. 2-15 2.2.4 RELATIONSHIP BETWEEN LATERAL AND VERTICAL SURFACE MOVEMENT Since the soil displacement is related to the mode of deformation at the excavation wall, it seems reasonable to assume that a consistent relation¬ ship exists between the lateral and vertical soil movements, depending on the type of wall deformation. Hence, it would be beneficial to examine soil be¬ havior under the influence of various wall distortions to gain an insight as to how the soil displacements are transmitted to the ground surface where buildings and other structures derive their support. Figure 2.7 summarizes the results of two model tests that were per¬ formed by Milligan (31) at Cambridge University. In both tests a flexible, smooth wall was used in combination with dense sand (rounded, coarse quartz with initial void ratios of 0.54 and 0.55). Soil strains and displacements were measured by means of X-ray radiographs that were taken of lead shot embedded in the soil matrix. As the soil deformed. X-ray pictures provided a direct measure of soil strain by showing the relative displacements of lead shot. The ground movements are shown at a scale of 1.5 times the actual measured displacements. All dimensions, as well as the lateral wall movement, are expressed in dimensionless form as a function of the maximum excavation depth. The wall deformations shown for the model tests are the results of a series of measurements taken at successively deeper excavation levels. For each stage of measured deformation, the patterns of soil movement remained un¬ changed and the magnitudes increased in proportion to the wall movements. Con¬ sequently, the final displacement vectors shown in Fig. 2.7 are representative of the movement patterns that developed for wall deformations as small as one-tenth those indicated. 2-16 Loterol pispl, Loterol Pis pi. Max. Depth of Excaw Max. Depth of Excav. Fig. 2.7 Soil Movements Related to Model Tests with Sand 2-17 The different types of wall defonnation are reflected in the pattern Of soil coyement. Hear the ground surface, the vector displacements for the cantilever movement of the wall show a ratio of lateral to vertical displace¬ ment ranging between 1.3 and 1.5. By way of contrast, the vector displace¬ ments for the Inward bulging of the wall show a more complicated pattern of orientation. Horizontal restraint at the upper level of excavation has limned the development of lateral movement at this elevation. Correspondingly. the ratio of lateral to vertical displacement near the ground surface ranges between 0.5 to 0.7. Figure 2.8 summarizes the ratios of the horizontal to vertical sur¬ face movements that were measured for both the model tests and braced excava¬ tions in Washington. 0. C. Ratios of horizontal to vertical displacement for the braced cuts were obtained from two different excavations. Settlement and inclinometer measurements, associated with the 60 ft (18.3 m) excavation described in the previous section, were combined to estimate ratios of surface movement within approximately 35 ft (10.7 m) of the edge of the cut. At greater dis¬ tances from the excavation, precise optical leveling and tape extensometer readings near the column footings of a highrise apartment building were used. The apartment building is located within 40 ft (12.2 m) of a 60 ft (18.3 m) excavation. The ratios correspond to a time when the excavations had been deepened to subgrade and the bottom-level braces had been removed. Both exca¬ vations were characterized by a similar soil profile and construction proce¬ dure. The distances from the edge of excavation, within which the ratios of horizontal to vertical displacement were estimated, are expressed as a frac¬ tion of the maximum excavation depth. Ratio Of Horizontal To Vertical Movement, dH/dV 2-18 Distonce From Edge Of Excavation Maximum Depth Of Excavation Fig. 2.8 Ratio of Horizontal to Vertical Ground Movements for Model Tests and Excavations in Washington, D. C. 2-19 The results of the model tests serve to bracket a range of surface displacement patterns that can be anticipated within the zone of plastic soil behavior. Within this framework the field data appear to represent a reason¬ able ratio of lateral to vertical displacement. The field data indicate that the ratio of horizontal to vertical movement increases with distance from the edge of excavation, ranging from the value of 0.7 at a distance of 0.25 the maximum excavation depth to nearly 1.0 at distances exceeding the maximum excavation depth. 2.3 GROUND MOVEMENTS RELATED TO BRACED EXCAVATION IN CHICAGO 2.3.1 SOIL CONDITIONS IN CHICAGO The City of Chicago is founded on a series of till sheets that were deposited during the Pleistocene Epoch. Much of the downtown or "Loop" area of Chicago is underlain by a stratum of soft, compressible clay that was deposited as part of this glacial sequence. All the Chicago cuts studied in this report were extended into the soft clay and thus, their performance is indicative of the relatively large, plastic deformations that accompany excavation in this type of material. Figure 2.9 shows a typical soil profile for downtown Chicago. Repre¬ sentative values of undrained shear strength are listed for the clays and typical standard penetration rates are listed for the granular soils. A notable stratum of soft clay occurs at a depth of approximately 15 ft (4.6 m) below the street surface. The soft clay grades into clay of medium consis¬ tency at a depth of approximately 40 ft (12.2 m), however the transition zone 2-20 Soil N Cu Fill Stiff Clay Soft Cloy Medium Cloy Stiff Clay Hard Clay Sand , Gravel, Boulders •77m'77T!!r-7rK!s~77^ Limestone Symbols! N - Standard Penetration Resistonce, Blows/ft Cu- Undrained Shear Strength, KSF (Pg x 10^) 15 To 25 1.5 (7.2) 0.4 ( 1.9) 0.75 (3.6) 1.5 (7.2) >4.0 (>I9.2) >70 — \J 20 ■( 6 . 1 ) 40 '( 12 . 2 ) 60 '( 18 , 3 ) 80 "( 24 . 4 ) _I00 "( 30 . 5 ) I i I I Fig. 2.9 Soil Profile for Braced Excavations in Chicago Depth, ft (in.) 2-21 may occur as much as 10 ft (3 m) higher or lower, depending on location. The soft clay is capped with a relatively thin layer of stiff, desiccated clay. A stratum of hard clay and silt, referred to as hardpan, occurs at a depth of approximately 65 ft (19.8 m). The depth of the limestone bedrock ranges from 95 to 110 ft (30 to 33.5 m). The water table is located at the top of the soft clay stratum. 2.3.2 SURFACE SETTLEMENT AND CONSTRUCTION METHOD The surface settlements associated with 9 different braced excavations are summarized according to the methods of construction in Table 2.1 and plotted in dimensionless form in Fig. 2.10. The settlements and distances are expressed as percentages of the maximum excavation depth. The settlements represent the final stages of the excavation sequence and, as such, correlate with the removal of the upper level braces during construction of the basement walls. With the exception of minor variations in the thicknesses of individual strata, the soil profile associated with the summarized excavations is essentially constant. Consequently, Fig. 2.10 allows for a comparative analysis of the soil movements as a function of the support conditions and construction technique. Three zones of ground displacement have been distinguished in Fig. 2.10 and related to the salient characteristics of construction. These zones approxi¬ mate the three zones of settlement delineated by Peck (36) with the exception that the widths of the settlement zones are noteably shorter than those indi¬ cated by Peck. As such, the settlements associated with the Chicago excava¬ tions are confined to areas that are comparatively nearer the edge of excavation. Peck’s summary, however, includes data from braced excavation in TABLE 2.1 INFORMATION RELATING TO CHICAGO EXCAVATIONS Cose 1 Symbol Depth H, ft (m) Wall Support Excavation Procedure Special Characteristics 1 O 44 (13.4) 14 BP 73 On 7 ft ,2.1 m) Centers With Lagging 2 Upper Levels of Cross-Lot Struts, Bottom Level Rakers; Upper .evels Preloaded; 12 ft 3.7 m) Vert. Space Excavate 14 ft (4.3 m) Below Previous Strut Excavate Center With Berm On Sides Of Cut Along Sooth Woll Where Increased Settlement Occurred Berms Removed And Bottom Level Rakers Instolled la 2 A 27 (8.2) Sheet Pile MZ 27 3 Raker Levels; 8 ft 2.4 m) Vert Space, Pre loaded Excavate Center With Berms Adjoining The Wall _ 3 □ 44 (I3l4) 30 in. (76.2 cm) Slurry Woll Upper Level Tiebacks, Bottom Level Rokers ; 16 ft (4.9m) Vert. Space; Pre loaded Excavate 14ft (4.3m) Install Tiebacks Excavate Center With Berms 4 7 30 (9.2) 10 HP 42 On 5 ft (1.5 m) Centers With Lagging 2 Raker Levels; 10 ft (3m) Vert. Space; ’re loaded Excavate Center With Berms Adjoining The Woll Insufficient Woll Support Due To Delay In Raker Installation 5 0 26 (7.9) 30 in. (76.2 cm) Slurry Wall 1 Strut Level No Prelood Excavate 15 ft (4.6m) Install Struts Excavate Center Lost Ground Associated With Caisson Construction 6 • 28 (8.5) 21 WF76 On 6.5ft (2m) Centers With Lagging Cantilever Support With Some Rakers Excavate 14 ft (4.3 m) Drill Caissons Excavote To Subgrade Lost Ground Associated With Caisson Construction And Insufficient Wall Support 7 7 45 (13.7) 30 in. (76.2cm) Slurry Wall 3 Raker Levels; 11 ft (3.4 m) Vert. Space ; Preloaded Excavate Center With Berms Adjoining The Wall Lost Ground Associoted With Coisson Construction 8 A 70 (21.3) Sheet Pile MZ 38 6 Strut Levels; 10 ft {3m) Vert. Space Preloaded — Excavate 12 ft (3.7m) Below Previous Strut 9 1 ■ 1 37 (11.3) Sheet Pile MZ 38 3 Raker Levels; 10 ft (3m) Vert. Space ; No Prelood Excavate Center With Berms Adjoining The Wolls Distance From Edge Of Exc ovotion Depth Of Excavation Zone I —Well Braced Excavations With Slurry Wall Or Substantial Berms Left Permanently In Place Zone II -Excavations With Temporary Berms And Raker Support Zone III-Excovations With Ground Loss From Caisson Construction Or Insufficient Wall Support Fig. 2.10 Summary of Settlements Adjacent to Braced Cuts in Chicago 2-23 2-22 2-24 Oslo (ll) where the depth of soft clay beneath the excavation bottom was, in most instances, considerably larger than for the Chicago cuts. The excavations summarized in Table 2.1 were braced cuts associated with deep basement construction. Typically, this type of opencutting is ex¬ tended an initial 6 to 8 ft (1.8 to 2.4 m) to provide a working level from which drilled caissons are installed. The central portion of the excavation then is deepened to subgrade as berms are left in place agains.t the excavation walls. Grade beams and basement slabs are constructed in the central area of the cut while wall support is provided by installing rakers between the com¬ pleted foundation and sheeting line. Several levels of rakers may be required, depending on the depth of cut. As the temporary berms are removed in stages, successively deeper levels of rakers are installed. Zone I includes settlement data associated with cases 1 and 3. In each case, the upper level supports were installed and preloaded before the central portion of excavation was deepened to subgrade. In case 1 substantial berms were left permanently in place along the north and west walls of the soldier pile lagging system. In case 3 the bottom levels of the slurry wall were supported by temporary berms until rakers were installed and preloaded. Zone II includes settlement data related primarily to excavations with temporary berms and raker support. The excavation support includes soldier pile-lagging walls and sheet pile walls. Where rakers were preloaded and in¬ stalled on small vertical spaces, as was the condition for case 2, settlements were restrained. Correspondingly, they plot near the upper portion of the zone. 2-25 The largest settlements measured for the Chicago excavations are shown in Zone III and are related either to ground loss from caisson con¬ struction or to insufficient support of the excavation wall. The data for cases 5 through 7 pertain to excavations where the drilled caissons were excavated without slurry. Consequently, squeezing ground caused by lack of restraint in the open holes, especially in the soft and medium clay strata, is responsible for part of the settlement. For case 7, additional lost ground during caisson installation can be attributed to excessive pumping of water and fines from the sand and gravel stratum overlying bedrock. It is unlikely that the settlements associated with Zones I and II were influenced significantly by caisson construction. In these cases when drilled caissons were installed, they were excavated under guidelines that called for minimizing the hole dimensions with respect to the temporary casings and maintaining a bentonite slurry during drilling. In addition, inclinometer measurements for the excavations in question did not indicate deep ground movements during caisson construction. A better understanding of the excavation procedure and associated ground movements can be obtained by examining the inclinometer measurements for several excavations that used different methods of wall support. Infor¬ mation of this nature is assembled in Figs. 2.11, 2.12, 2.13 and 2.15 where lateral displacements for each of several cuts are illustrated in combination with the soil profile and a scale representation of the excavation levels. Where possible, settlement profiles are shown in relation to the lateral wall movements. The dates corresponding to the soil displacements and levels of excavation are referenced to the beginning of caisson construction. \ \ 2-26 Lateral Displacement, in. (cm) 20 (6.1) 40 ( 12 . 2 ) 60 (18.3) 3 (O I k 55 » o a> m 80 (24.4) o. a 0 '' E 20 « o (6.1) O r 3 (/) 40 % (12.2) g 55 9 60 .2 (18.3) 2 80 (24.4) a « o Fig. 2.12 Soil Displacements for Case la 2-27 Figures 2.11 and 2.12 show the lateral wall displacements associated with case 1 and case la, respectively. The wall of this excavation was com¬ posed of 14BP73 soldier piles on 7-ft {2,1-m) centers with wood lagging installed between adjacent soldier piles. The wall was supported in its upper levels by two tiers of cross-lot and diagonal struts that were preloaded to 50 percent of the design brace load. The central portion of the excavation was deepened to a subgrade level of 44 ft (13.4 m) below the surrounding street surface. Large berms were left in place against the north and east walls of the cut as shown in Fig. 2.11. The berms were approximately 25 ft (7.6 m) wide at the top and were sloped at an angle of roughly 30° from the horizontal. The lateral wall displacements and surface settlements before and after excavation of the central area of the cut are indicated in the figure. By way of contrast. Fig. 2.12 shows the lateral displacements for the south wall of the excavation where the berms were removed and replaced with raker supports. A comparison of the two figures indicates that the transfer of lateral restraint from berms to the raker sys¬ tem was relatively inefficient. Installation of the bottom rakers required that the berms be diminished to a size appropriate for insertion of the raker support. Excavation of this nature decreases the lateral restraint at the excavation wall as well as reduces the dead weight of soil acting to limit bottom heave. The reduction of berm support in combination with deforma¬ tion and adjustment of the rakers contributed to the prominent inward bulging shown in Fig. 2.12. The increased displacement emphasizes the need for care¬ ful excavation of the soft clay in combination with prompt installation of stiff bracing. 2-28 Lateral Displacement, in. (cm) (I9i2) (IQ2) (5.0) Fill Stiff Clov Soft to Medium Clay - Stiff to Very Stiff Clay - Hard Clay and Silt - 20 (6.1) 40 ( 12 . 2 ) 60 KI8.3) 80 (24.4) Distance From Edge of Cut,ft (m) Fig. 2.13 Soil Displacements for Case 9 Fig. 2.14 Wall Movement Associated with Lateral Displacement of Caissons Depth Below Street Surface, ft (m) 2-29 Figure 2.13 shows the lateral wall displacements associated with case 9. The wall of this excavation was composed of MZ 38 sheet piles that were supported by three levels of rakers. The rakers were not preloaded during installation. Lateral movement and settlement profiles are shown for excavation levels corresponding to a depth of 13 and 37 ft (4 and 11.3 m). Most of the lateral displacement developed in the strata of soft to medium clay in a manner similar to the ground deformations shown in Fig. 2.12 As in Fig. 2.11, the volume of lateral wall movement is approximately equal to the volume of settlement behind the sheeting line. Substantial displacements occurred as both the excavation was deepened to subgrade and the temporary berms were removed during raker installation. Large increases in lateral de¬ formations, ranging from 4 to 5 in. (10.2 to 12.7 cm), occurred at all raker levels as the excavation was extended into the soft and medium clays. In this type of excavation the rakers transmit their loads to the completed portion of the foundation. Correspondingly, most of the earth pres¬ sures generated at the wall of the cut are balanced by 1) the lateral resistance of the caissons and 2) the adhesion between the bottom soils and the basement slabs. Commonly, the grade beams for a given foundation are constructed several weeks in advance of pouring the basement slabs. Since the grade beams are connected to the caissons, rakers that are braced against the grade beams transmit the greatest portion of their load to the caissons. The corresponding movement of the caissons and rakers is shown in an exaggerated form in Fig. 2.14. The elastic displacement of the caissons can be estimated with the aid of the dimensionless charts described by Davisson and Gill (9) and Davisson (8). 2-30 Estimates performed on this basis reveal that, if only 25 percent of the anticipated earth pressure for a 40-ft (12.2-m) cut in Chicago soils is trans¬ mitted to a grade beam connecting two caissons, lateral movements between 1 and 3 in. (2.5 and 7.6 cm) can develop for caisson diameters of 5 and 3.5 ft (1.5 and 1.1 m), respectively. Lateral movements of the caissons diminish the effective raker stiffness and cause displacement at the level of support on the excavation wall. Consequently, optimal bracing requires that raker in¬ stallation be coordinated with the construction so that raker loads can be transmitted to a suitable foundation bearing. In addition, preloading the rakers increases the effective support stiffness by taking up the initial separations in the bracing line and promoting a flush contact between the caissons, grade beams, and basement slabs in the area of raker abutment. Figure 2.15 shows the lateral wall displacements for case 7. The large ground movements, which occurred during caisson construction at this site have not been indicated in order to concentrate on the specific dis¬ placements that developed during opencutting. Hence, the lateral displace¬ ments have been referenced to a time after the completion of caisson construc¬ tion and before excavation in the soft and medium clays. The excavation was supported by a 30-in. (76.2-cm)-thick, concrete slurry wall, restrained at three levels by rakers. The upper two raker levels were preloaded to 50 percent of the design load. The installation of the first level rakers occurred while substantial berms were in place. The installation of all raker levels corresponded closely with construction of the basement slabs in the areas of raker abutment so that a sound foundation bearing was provided. The wall displacements extend to a deeper level than those shown on Figs. 2-31 Lateral Displacement, in. (cm) 0 (IS.2) (10.2) (5.0) 6 4 2 * 8 ■ T " -r ? I ■ 1 Fill Stiff Cloy Soft to Medium Cloy Stiff Cloy Hard Clay and Silt 20 ( 6 . 1 ) 40 ( 12 . 2 ) 60 (18.3) 80 (24.4) Distonce From Edge Of Cut Depth Of Cut " 1-0 ns 2.5 3 - g X > A C O o £ O w O 3 o> 9 - IB - 21 *- Zone For Chicogo Braced Cuts in Soft To Medium Cloy Depth ; 27 (8.2) To 44 ft (13.4 m) Case Symbol Max. Depth,ft (m) r O 44 (13.4) la 44 (13.4) 2 L 27 (8.2) 9 ■ 37 (11.3) \ Fig. 2.16 Angular Distortion for Braced Cuts in Chicago Depth Below Street Surface, ft (m) 2-32 2.13 and 2.14, and reflect the nature of the concrete wall whose stiff section tended to transmit movement into the underlying soils. The volume of lost ground is approximately one-half that for the sheet pile excavation shown in Fig. 2.13 that used a similar scheme of raker support. The volume of measured wall displacements for case 7 compares favorably with the volume of slurry wall movements for case 3 that have been summarized in detail by Gnaedinger, et al., (14). The relatively small volumes of movement associated with deep excavation in both these cases indicate that slurry walls, when used in combination with excavation control and careful installation of braces, can result in relatively small soil displacements. Figure 2.16 shows the angular distortion associated with braced exca¬ vations in Chicago plotted as a function of the dimensionless distance from the edge of excavation. The angular distortion was estimated in a manner similar to the method used for the braced excavations in dense sand and interbedded stiff clay. The values of angular distortion shown in the figure are derived from the measurements associated with the excavations listed as cases 1, la, 2 and 9 in Table 2.2. Settlement profiles were available for these cases only. On the basis of the data indicated, there is no consistent relationship between excavation depth and angular distortion. Angular distortion appears to be smallest for case 1 where permanent berms were left in place. Ground loss associated with caisson construction was not significant for these excavations and thus, the angular distortions are related directly to opencutting. The recommended zone of angular distortion has been delineated on the basis of settlement data within a range of distances from 0.5 to 1.5 times the maximum depth of excavation. 2-33 2.3.3 LATERAL GROUND MOVEMENTS The lateral surface movements associated with 4 different braced excavations are summarized in Fig. 2.17. These excavations represent the cases where both ground loss from caisson construction was not significant and lateral survey data were available. The lateral displacements and dis¬ tances are expressed as percentages of the maximum excavation depth. The lack of sufficient excavation support is indicated by the relatively large move¬ ments associated with case 4. Because the data is limited, it is difficult to recommend zones of lateral movement. The estimated zone for good to average workmanship is based primarily on the data from cases 1 through 3 where the corresponding distances from excavation are concentrated between 0.30 and 0.75 times the maximum excavation depth. 2.3.4 THE RELATIONSHIP BETWEEN LATERAL AND VERTICAL SURFACE MOVEMENT The relationship between lateral and vertical surface displacement is closely associated with the mode of deformation at the excavation wall. For excavation in sand, it has been shown that wall deformation related to canti¬ lever movement and inward bulging each result in a characteristic ratio between the lateral and vertical surface displacement. In a similar fashion, it is useful to examine the ratio of lateral to vertical soil displacement for braced cuts in Chicago. If a characteristic ratio can be shown as a function of the wall deformation, then this relationship would form a basis for estimating the lateral distortions associated with opencutting where lateral displacement data is limited or nonexistent. The relationship between horizontal and vertical ground movements has been studied for five braced cuts in Chicago. In all cases, ground loss \ La teral Displacement Maximum Depth Of Excavation ’ 2-34 Dietance From Edge Of Excovotion Maximum Depth Of Excavation e2 gso 0.5 t.o A A A I o ° 0° i j7-cr 1.00 - \ - 7 ^ 7 77 7 7 7 7 1.5 7 7 7 1.50 2.00 230 -1 Estimated Zone For Temporary Berm And Raker Excavation In Soft To Medium Cloy ; Good To Average Workmanship Case Symbol 1 O 2 A 3 □ 4 7 (See Table 2.1) Fig. 2.17 Summary of Lateral Displacements Adjacent to Braced Cuts In Chicago Loteroi Displacement, in. (cm) Average = 0.74 Standard Deviotion =0.35 (S.0) (10.2) (15.2) Fig. 2.18 Ratios of Horizontal to Vertical Ground Movement for Case 1 Depth, ft (m) 2-35 associated with caisson construction was not significant and hence, the data are representative of soil movements that are related directly to opencutting. All the data have been screened in accordance with two guidelines: 1) Each ratio of lateral to vertical displacement derives from the combined measure¬ ment of lateral movement and settlement at the same point, 2) Measurements equal to or less than 1/4 in. (0.6 m) have been neglected in order to minimize the influence of survey error on the computed ratio. Information pertaining to the ratio of horizontal to vertical surface displacement is summarized in Figs. 2.18 through 2.22. Each figure provides a graphical representation of the data distribution in the form of a histogram. The average ratio of the horizontal to vertical displacement, as well as the standard deviation, is indicated. A lateral displacement profile is presented in each figure that shows the wall deformation corresponding to the ground movements from which the ratios of the horizontal and vertical ground displace¬ ment were calculated. Figures 2.18 and 2.19 show that the average ratios of horizontal to vertical movement are 0.74 to 0.60 for cases 1 and 2, respectively. The lateral wall displacements for these excavations show substantial inward bulging and thus the low ratios reflect horizontal restraint in the upper levels of the excavation. By way of contrast. Fig. 2.20 shows an average ratio of horizontal to vertical movement of 1.68 for case 3 where cantilever deformation of the exca¬ vation wall is apparent. Figures 2.21 and 2.22 show prominent cantilever defor¬ mation with a slight Inward bulging of the excavation walls for cases 4 and 10*, * The excavation associated with Case 10 is not listed in Table 2.1, having been only partially completed at the time of writing this report. 2-36 Ratio of Horizontal To Vertical Movement, dH/dV Averoge = 0.60 Standard Deviation - 0.08 Lateral Displacement, in. (cm) (^) (102) (l|2) I I r" r"i " i " i E ♦- Q. a> O Fig. 2.19 Ratios of Horizontal to Vertical Ground Movement for Case 2 Lateral Displacement, in. (cm) (90) oaz) (19.2) 20 ( 6 . 1 ) 40 ( 12 . 2 ) 60 (IS.S) 60 (24.4) Fig. 2.20 Ratios of Horizontal to Vertical Ground Movement for Case 3 Depth, ft (m) k 2-37 Ratio of Horizontal To Vertical Movement, dH/dV Averoge = 1.32 Standard Deviation =0.50 Lateral Displacement, in. (cm) tS.0) (102) (ISi) (20.3) ° ^ ^ 6 a r— I I —I— r— r-i Fig. 2.21 Ratios of Horizontal to Vertical Ground Movement for Case 4 Lateral Displacement, in. (cm) (5.0) (10.2) (19.2) 0 2 4 6 I—I I I I —ri Rotio of Horizontal To Vertical Movement, dH/dV 0 20 ( 6 . 1 ) 40 ( 122 ) Averagt =1.37 Stondard Devlotion =0.57 Fig. 2.22 Ratios of Horizontal to Vertical Ground Movement for Case 10 Depth, ft (m) ^ , Depth, ft (m) 2-38 respectively. The raties of horUoetal to vertical displacement for these excavations range from 1.32 to 1.37 and reflect the comMned Influence of bulging and cantilever wall movements. The information su-arlted In Figs. 2.18 through 2.22 related to surface measurements that were taken within a distance of .35 to 1.0 times the maximum excavation depth from the edge of the cut. The scatter of the data and the limited range of distances prevent the delineation of a clear relationship between the ratio of lateral to vertical movement and distance from the excavation. To illustrate the influence of wall deformation on soil movements, a coefficient of deformation. is defined in Fig. 2.23. The numerator uf the term is a measure of the cantilever portion of wall movement and is expressed as the lateral displacement, a, at the top brace level, correspondingly, the denominator is a measure of the inward bulging and is defined on the basis of the displacement, a’, separating the point of maximum bulging from the line of rigid wall rotation. The coefficient of deformation indicates the relative amounts of cantilever movement and inward bulging that are developed at the excavation wall. Consequently, it can be related to the pattern of ground movement behind the sheeting line. The average ratios of horizontal to vertical displacement for the five excavations, referenced in Figs. 2.18 through 2.22, are plotted as a function of the coefficient of deformation in Fig. 2.23. For values of the coefficient greater than 4 and less than 1, the ratio of lateral to vertical displacement Ratio Of Horiz. To Vertical 0 i s placemen t, d H/d V 2-39 ig. 2.23 Relationship Between the Ratio of_Hor1zontalJ0^Vertica1 2-40 approaches 1.6 and 0.6, respectively. It should be emphasized that the coef¬ ficient of deformation is only a rough gaging of the wall distortion and will not be simple to apply to profiles of wall movement that have been influenced by high preloading of the braces. However, within the limits of interfer¬ ence from preloading, the coefficient of deformation can be a tool for relating the shape of the deformed wall with the pattern of surface dis¬ placements behind the sheeting line. In this way, the surface displacements can be related to the excavation method. For example, if the upper level supports are installed without adequate stiffness or berms are cut back substantially before raker installation, cantilever movements will predominate and lateral displacements will develop in excess of settlement. By way of contrast, if the upper level supports are installed early in the excavation program in such a manner that they have sufficient stiffness, lateral surface movements will be restrained to values less than those of the settlements. 2.4 SUWIARY The ground displacements associated with recent excavation in Washington, D. C. and Chicago have been examined in light of the soil profile and construction methods in each area. The typical surface settlements and lateral displacements for excavation in both cities have been summarized in dimensionless form as a percentage of the maximum excavation depth and in terms of angular and lateral distortion. The ratio of horizontal to vertical sur¬ face displacement has been examined and shown to be useful for two reasons: 1) the ratio of horizontal to vertical movement is related to the deformed li 2-41 shape of the excavation wall and is diagnostic of certain excavation methods, 2) measurements of lateral movement are usually scarce and hence, typical ratios of horizontal to vertical surface movement can be used to estimate lateral displacement from settlement data. On the basis of the assembled diagrams and corresponding information, zones of angular and lateral distortion associated with various excavation pro¬ cedures are described in Table 2.2. Distances from the edge of excavation are expressed as a fraction of the maximum excavation depth. The values of angular and lateral distortion associated with excavation in dense sand and interbedded stiff clay have been determined from measurements related to opencutting in Washington, D. C. For these cuts, the ratio of horizontal to vertical surface movement was estimated as approximately 0.8 within a distance from the exca¬ vation equal to one-half the maximum excavation depth. The values of angular and lateral distortion associated with excavation in soft clay have been determined from measurements related to opencutting in Chicago. For these cuts, the ratio of horizontal to vertical surface movement was estimated as approximately 1.3 within a distance from the excavation equal to the maximum depth of excavation. The difference between the ratios for cuts in dense sand and cuts in soft clay is primarily a function of the excavation technique. The maximum wall movements associated with excavation in Washington, D. C. occurred in the deepest portions of the cuts when the upper levels were restrained with cross¬ lot braces. By way of contrast, the temporary berm and raker construction used in the large building excavations of Chicago often resulted in sub¬ stantial cantilever movement before the upper rakers were installed. This was / 7^ 7 -* 0 0 3“ tn 0 tn 0 rt rr rr r+ O 0 0 0 (V X ro 0 n • • • • Of tn tn 0 tn < Of rT 0 3 o o o o rt O ro tn ro cn o 4^ O rt- O ro CL n> T3 r+ (jl vO t/i ''O ^ ui cn r+ O O O O O <-n r\> Ln C3 Crt r+ 3 O -5 O 3 a> Cl id a> a> X o o; < Qt O 3 <—«. t/> 3> X 3 O CO —• c O r+ —• I -*• ft» OJ O -5 — 3 O ^ irt r“ X r*- 0> O ri* -J -5 a> O c+ -5 I U) O —• ^ 3 0 CL O' m ft) X tn > -0 -% 0 tn to ro tn X c* 3 rt 3 Of 0 UD 3“ < ct <-r rT rr -5 c •» Of O) 0 0 0 0 0 —J 3 rr 1 Q. -»• to 0 -5 ^ ro 0 0 tn to ro —' 3 tn T 3 a> x* -*• 07 0 (T> 3 < rt- 0 4 :* tn tn 0 C -h -h •3 rr "O —*• 0 n ^ trt |— X rt a> O rf —* fD O <-► -J I a* OJ o —• •W* 3 o> cc CT> —* -*• m O O 3 X c+ O -h fO 0> r+ t/> -5 < <-► cr ft» H TfBrt -ICQ.-'* 00 Q. O • (/) 0) 3 CO-. CL 3 (/) 3 ^■O ri* -O Q. “5 -t> fD O -h 3 X CO -»• n Q 3 -• a> (o rt*^ rD 3 CL CL n (V ^ ot *0 0 3 <-♦• to Q. 3* CO • • I f+ O* a ”o <-► *o o 3 3 o -o X o -*• -5 3 Of Of -5 O > (✓5 CO O o HH ■H m o 03 ? o m X o < o z 00 2t7-2 ZONES OF ANGULAR AND LATERAL DISTORTION 2-43 not always the case, however, as is evidenced by the ground movements related to cases 1 and 2 where the upper braces were installed early in the excavation program. Hence, the lateral distortions shown in Table 2.2 for cuts in soft clay could be decreased by 25 to 50 percent if the excavation is designed suc*i that the wall will be braced adequately while substantial berms are still in place. Furthermore, slurry walls in combination with conscientious excavation and bracing have resulted in relatively small displacements. Field observations (10,19,41) suggest that the settlement profiles adjacent to slurry walls tend to be concave upward as opposed to the downward turning profiles shown by other Excavation schemes, especially those using soldier piles and lagging (34,35). It seems reasonable, therefore, that the angular and lateral distortions indi¬ cated in Table 2.2 should be diminished if applied to a well braced, slurry wall system. At present, additional measurements and study are needed to develop a clear picture of the ground movements associated with slurry wall construction. Angular distortion provides a measure of the change in shape or slope, of the settlement profile. The deformed shape of the ground surface can then be related, by direct analogy, to the deformed shape of a structure. However, the degree of correspondence between the slope of the ground sur¬ face and the deformed shape of the structure is a function of the geometry of the structure relative to the curved portion of the settlement profile and the stiffness of the structure relative to the stiffness of the foundation soil. In a similar manner, the lateral distortion is a measure of the lateral strains imposed on a given building. However, the degree to which the building reflects the imposed ground strains is a function of the stiffness of the building and of the forces mobilized between the building and the deforming soil. Although \ 2-4 4 angular and lateral distortions of the soil provide a first estimate of building deformation, judgments pertaining to structural response, especially potential instability, can be made only by studying the specific buildings. Consequently, it is useful to regard the values of angular and lateral dis¬ tortion summarized in this chapter as the upper limits of strain imposed on a given structure, realizing that further examination of the structure may be required to determine its stiffness and construction details. 3-1 CHAPTER 3 ARCHITECTURAL DAMAGE TO BUILDINGS 3.1 INTRODUCTION 3.1.1 DEFINITIONS OF DAMAGE Building damage generally is divided into two categories: 1) Architectural Damage, i.e., damage that pertains to the cracks or separations in panel walls, floors, and finishes. 2) Structural Damage, i.e., damage that relates to the cracks or distortions in primary support elements such as beams and columns. Frequently, a further distinction is made on the basis of the building services. This type of damage, referred to as functional damage, pertains to damage that impairs the use of the structure. The present chapter deals solely with architectural damage. In this chapter, architectural damage is studied by summarizing previous re¬ search, developing correlations between architectural damage and differential movement on the basis of field evidence, and discussing the influence of archi¬ tectural damage on the use of various buildings. 3.1.2 CORRELATION OF ARCHITECTURAL DAMAGE WITH DIFFERENTIAL MOVEMENT Two parameters are commonly used for developing correlations between architectural damage and differential settlement. These parameters are the angular distortion and the deflection ratio. As defined in the previous chapter, angular distortion, 5^, is the differential settlement between two points divided by the distance separating them. When related to building damage, angular distortion is coimonly modified by subtracting \ 3-2 the rigid body tilt from the measured settlement. In this way the modified value is more representative of the deformed shape of the structure. The deflection ratio, hjl, is defined as the maximum displacement, A, relative to a straight line between two points divided by the distance, 1. separating the points. Both parameters are illustrated in Fig. 3.1 as they apply to the settlement of a building adjacent to excavation. The deformed shape of the building is exaggerated for purposes of illustration. The shape of the settle¬ ment profile is convex and is characterized by a slope that increases with diminishing distance from the excavation. Field data (34,35) indicate that this type of profile represents the general condition for many braced cuts. In the figure, rigid body tilt is indicated by the angle, a. Often, rigid body tilt is extremely difficult to estimate on the basis of settlement, especially settlement related to opencutting. In addition, the settlement profile associated with opencutting develops in stages as the excavation is carried to subgrade. Consequently, the building is subjected to a settlement wave that causes bending and shear distortion in the structure even though the final slope of the settlement profile may be constant. In most cases, the building dimension perpendicular to the edge of excavation is relatively large with respect to the length of the settlement profile. For the general case, therefore, the evaluation of angular distortion directly from differential settlement is a reasonable way to estimate the deformed shape of the building. Rigid body tilt, however, must be considered in cer¬ tain applications, especially when judging the influence of settlement on stiff, narrow structures. In these cases, rigid rotations should be evaluated 3-3 Deflection Ratio, A/J L/Jl * Fig. 3.1 Building Deformation Caused by Settlement in Response to Opencutting Fig. 3.2 Building Deformation Caused by Lateral Displacement in Response to Opencutting \ \ \ \ \ 3-4 from inclination measurements of the front and rear walls, in addition to settlement data. No single expression for differential settlement is clearly the most expedient for correlation with damage. Both parameters possess advan^- tages that make them useful for different applications. For example, the deflection ratio is advantageous in that: 1) it is closely related to the radius of curvature and, hence, a good indicator of bending deformation, and 2) it provides a direct measure of the deviation from uniform settlement and rigid body tilt. Polshin and Tokar (38) and Borland and Wroth (6) have used the deflection ratio as a convenient index for studying the damage sustained by continuous bearing walls that have settled over their full length. For a building adjacent to opencutting, however, the curved portion of the settle¬ ment profile frequently develops such that the ratio of the deformed building length to building height (L/H) is approximately equal to or less than unity. Shear strain, for this geometric condition, contributes most prominently to the building deformation. Consequently, angular distortion is a useful parameter because it is directly related to shear strain. When applied to frame structures, angular distortion also provides a convenient means of relating the settlement of adjacent columns to the strains imposed in a structural bay. Furthermore, correlations based on angular distortion, by virtue of their simple definition, can be related to a broad range of field observations and compared directly with previous research. In this chapter, both angular distortion and the deflection ratio are discussed in the context of previous criteria for architectural damage. Only angular distortion, however, is used for correlating observed damage with differential building settlement caused by adjacent excavation. \ 3-5 Buildings near braced cuts also are influenced by lateral dis¬ placements. As was indicated in the previous chapter, the lateral soil strains associated with opencutting can be large and can represent a sub¬ stantial portion of the strain sustained by adjacent structures. Hence, it is important to judge the limits of building disturbance in light of the lateral ground distortion. Lateral distortion, 6^. is defined as the differ¬ ential lateral movement between two points divided by the distance separating them. Figure 3.2 shows lateral distortion as it applies to the horizontal movement of a building adjacent to excavation. The deformed shape of the building is exaggerated for purposes of illustration. 3.2 BASIC CONSIDERATIONS OF ARCHITECTURAL DAMAGE 3.2.1 PREVIOUS STUDIES OF ARCHITECTURAL DAMAGE Correlations between differential settlement and architectural damage have been the subject of extensive research. Various methods of analysis have been followed and different criteria for the first appearance of damage have been proposed. It is useful, therefore, to briefly suimarize the results of previous research as a baseline from which to extend the study of architectural damage to buildings influenced by both vertical and lateral movement. Essentially, two different methods of formulating damage criteria have been used. 1) Empirical correlations of architectural damage and differ¬ ential settlement have been developed on a statistical basis by analyzing the settlement data for a large number of buildings, 2) Theoretical models have 3-6 been developed by considering the critical tensile strain for building materials and the ratio of the deformed length to height (L/H) of the structure. These different methods and the work performed by the major proponents of each are summarized under the following two headings: Empirical Correlations : Skempton and MacDonald (40) reviewed the settlement histories of 98 buildings to set deformation criteria for damage to structures. The criteria indicate that cracks in panel walls of frame buildings or walls in load bearing wall structures are likely to occur if the angular distortion exceeds 3.3 x 10‘^ (1/300). These criteria were corroborated by Grant, et al., (16) in a more recent study of the settlements associated with 95 additional buildings. Furthermore, Grant, et al., recom¬ mended a threshold for architectural damage on the basis of the deflection ratio. According to their study, cracks in panel walls of frame buildings and walls of load-bearing structures are likely to occur if the deflection ratio exceeds 1.0 x 10'^ (1/1000). The broad background, on which the empirical correlations are based, include data from many areas and a great variety of observations. Although they represent some instances of subjective judgment, they, nevertheless, correspond to a substantial body of field evidence and, most importantly, reflect the perceptions of those who used the buildings. Theoretical Models : The theoretical models are based on two uni¬ fying concepts: , _ 1) Damage occurs when the instantaneous increase in building strain exceeds the critical tensile strain of concrete or brick masonry. Critical tensile strain is defined as the strain at which local fracture becomes visible. 3-7 2) The strain imposed on a given building is related to the geometry of the building as expressed by the ratio of the deformed length to height of the structure (L/H). Polshin and Tokar (38) were the first to formulate damage criteria on the basis of the above concepts. They developed an approximate theoretical relationship that predicted cracking in brick-bearing walls as a function of the critical tensile strain for brick masonry (0.05 percent), the deformed shape of the wall, and the ratio of the deformed length to height (L/H) of the wall. They compared their theoretical relationship with observations of several different structures. For L/H < 3, they specified limiting deflection ratios of 0.3 x 10'^ (1/3300) and 0.4 x 10'^ (1/2500) for buildings on sand and soft clay, respectively. In addition, they recommended an angular distortion of 2.0 x 10 ^ (1/500) as a threshold value for the appearance of cracks associated with in-filled framed structures. Borland and Wroth (6) extended and refined the work of Polshin and Tokar by developing models that predicted the onset of cracking for both bending and shear-induced deformation. In addition, they called attention to hogging, i.e., convex curvature, which is the principal mode of building deformation caused by opencutting and tunneling. Their models show that, for deformation geometries similar to those of buildings adjoining braced cuts (L/H < 2), cracks can occur in bearing wall structures at deflection ratios as low as 0.4 x 10’^ (1/2500). The criteria for initial cracking in response to differential settle¬ ment are summarized in Table 3.1. These recommendations represent the range of critical values generally encountered and, hence, are indicative of the field observations and recommendations of others (6,15,17). For in-filled frames, the criteria for initial cracking are in general agreement. For 3-8 TABLE 3.1 SUMMARY OF DAMAGE CRITERIA FOR INITIAL CRACKING IN BUILDINGS SUBJECT TO SETTLEMENT UNDER THEIR OWN WEIGHT Deformation corresponding to initial cracking In-filled frames Load bearing walls Angular Source distortion, 6^ Relative Angular deflection, distortion, 6^ Field observations .3 - Skempton & MacDonald (40) 3.3 x 10 3.3 X 10"^ Field observations _3 - Grant, et al. (16) 3.3 x 10 1.0 X 10"^ 3.3 X 10'^ Theoretical model and field observations _3 - Polshin & Tokar (38) 2.0 x 10 ^0.3 to 0.4 X 10'^ Theoretical model - Burland & Wroth (6) ^0.4 X 10'^ 1 L/H < 3 ^ l/W < 2 3-9 load-bearing walls, however, critical values of the deflection ratio are appreciably different. Deflection ratios based on consideration of critical tensile strain (Polshin and Tokar, Burland and Wroth) tend to support values almost three times lower than the values based on summaries of building settlement (Grant, et al.). 3.2.2 NOTICEABLE DAMAGE Deformation exceeding the critical tensile strain may cause local fractures that are spotted through the use of careful and well- directed observation, but may not necessarily be noticed by building occupants. For example, Littlejohn (26) has reported on the initial cracking of brick walls subject to mining subsidence. Cracks were detected at a deflection ratio of 0.16 x 10"^ (1/6130) and a lateral ■3 strain in the brickwork of approximately 0.25 x 10 . However, the cracks were observed as part of a systematic surveillance program and were only 0.004 to 0.01 in. (0.10 to 0.26 mm) wide. In a similar manner, a building adjacent to deep opencutting was instrumented as part of a measurement program performed by the University of Illinois. In the area of instrumentation, a 9-in. (0.23-m) thick, reinforced concrete wall was moni¬ tored for signs of distress. Small cracks, less than 1/64 in. (0.4 mm) wide. 3-10 were first noticed in the upper portion of the wall at an angular distortion of 0.3 X 10"^ (1/3300) and a lateral strain of 0.3 x 10 However, no cracks or separations were reported in other portions of the structure even though the majority of walls were finished with plaster and building personnel had been alerted to the possibility of such disturbance. Clearly, the criteria for architectural damage should reflect the dimensions at which cracks become noticeable, especially if applied to large excavation projects where movements associated with opencutting and tunneling are inevitable. The extent to which a crack becomes noticeable is a function of the surface on which the crack appears. This includes the location of the surface, its texture, and the ambient lighting. These characteristics can be illus¬ trated by reference to Figs. 3.3 and 3.4, which are photographs of a crack that developed in an in-filled partition wall of a steel frame building. The build¬ ing was located adjacent to a deep open cut where, in several structural bays, soil movements caused architectural damage similar to the crack shown in the figures. The partition wall was constructed of concrete hollow blocks that were finished with plaster on one side only. Figure 3.3 shows the crack, which was 1/32 in. (0.8 nm) wide, as it occurred on the unfinished side of the wall. The crack follows the mortar joints, where it is obscured by the texture of the concrete and delineation of the blocks. By way of contrast. Fig. 3.4 shows the crack as it occurred in the plaster finish on the opposite side of the wall. The crack reflects the stair-step pattern of the hollow blocks. Although it was still 1/32 in. (0.8 mm) wide, it is much more noticeable owing to the contrasting background and en-echelon pattern of propagation. 3-11 Fig. 3.3 Crack in a Hollow Block Wall Caused by Adjacent Excavation Fig. 3.4 Crack in the Plaster Finish of a Hollow Block Wall Caused by Adjacent Excavation 3-12 In this report architectural damage is distinguished as cracks in plaster walls that are equal to or greater in width than 1/64 in. (0.4 nri) and cracks in hollow block, brick, and rough concrete walls that are equal to or greater in width than 1/32 in. (0.8 m). Separations in tile floors are assumed to be evident at widths of 1/16 in. (1.6 mm). On this basis, the correlations coincide with reasonable limits of visible disturbance and should represent a threshold where distortions are noticed and reported by building occupants. 3,2.3 TILTING OF BUILDINGS Buildings subject to excavation movements often experience tilt, especially along the building line closest to the excavation. Visual effects associated with tilt may impose a limit on tolerable building distortion since movement out of plumb may become obdectlonable, even before cracks and dis¬ tortions are noticed. It is very difficult to set criteria for objectionable tilt because tolerance for this type of displacement depends on the use of the building and the nature of its environment. Occasionally, bnck-bearing wall structures contain facade walls that are as much as 2 and 3 in. C5.1 and 7.6 cmlout of plumb over a three to four story building height. These inclina¬ tions are tolerated by building occupants, who generally are not aware that floors are locally out of level or that there is an apparent lean to the building. For example, during construction of the Chicago subway, litigation was under¬ taken by store owners who claimed that adjacent excavation had caused tilting along their building fronts. Reference to a pre-construction survey, however, showed that the buildings had been previously out of plumb and that, in some 3-13 instances, ground movements had actually improved tfie verticality of the structures. Frequently, construction surveys of adjacent property only in¬ clude optical leveling, whereas it would be advantageous to incorporate addi¬ tional measurements of building inclination. Such measurements, which can be easily taken on corner structures by means of a transit, reference the building condition before construction and provide an ongoing measure of deformation against which complaints and potential problems can be evaluated. A rigorous treatment of tilt would require measurements and dis¬ cussion of building inclinations that are commonly accepted by the occupants in various neighborhoods and for various structures. In addition, serious scrutiny would have to be directed to tall buildings where relatively small inclinations tend to be emphasized by the great heights. This kind of study is outside the scope of this report and.correspondingly, special considera¬ tion of tilt is omitted from the treatment of architectural damage. 3.2.4 ADDITIONAL CONSIDERATIONS There are additional problems associated with developing criteria for architectural damage. For each structure, the limits of tolerable distor¬ tion are a function of the age and deterioration of the building. In fact, each building may be thought of as possessing a "strain memory" wherein the strains related to settlement under its own weight, structural modifications, nearby construction, and gradual deterioration with age are accumulated. Furthermore, a particular building may possess local areas of weakness where even small ground strains can be concentrated to cause observable damage. Kerisel (20) has conmented that the most prominent separations that occur in \, 3-14 deformed structures often appear at the junction of two walls of different rigidities. This general aspect of building damage can be illustrated by reference to Fig. 3.5. Figure 3.5 shows a plan view of a 10-story building with respect to the braced excavations that were performed on two sides of the structure. The building is composed of a reinforced concrete frame with two basement levels and a 4.5-ft (1.3-m) - thick reinforced concrete mat foundation. Prior to excavation, the building was underpinned either with continuous pit piers or pipe piles whose locations are shown by the shaded areas in the figure. Typical settlements, measured along the exterior walls of the build¬ ing, are indicated. The maximum angular distortion, calculated on the basis of available settlement data, is 0.8 x 10’^ (1/1250). Although settlement of the structure was relatively small, conspicuous damage occurred in the form of cracks in plaster walls and fallen ceiling tiles. This damage was re¬ stricted to the area of the expansion joint located 110 ft (33.6 m) from the closest excavation. On the roof, separation of the expansion joint was re¬ ported to be 1/2 in. (1.3 cm) in excess of the joint separations on lower floors. The concentration of damage indicates that the building strains were transmitted to the area of the joint where lateral movement and rigid body rotation of the building were reflected in local cracks and separations. Since the capacity of each building to tolerate strain is a function of its specific construction, previous strain history, age, and deterioration, the collection and summary of field data should not be regarded as setting definite limits on the appearance of architectural damage. Rather, correla¬ tions of observed damage and measured displacements specify a range wherein cracks and separations are most likely to occur and be noticed. 3-15 Fig. 3.5 Plan View of a 10-Story, Concrete Frame Structure with Settlement Caused by Adjacent Excavations 3-16 3.3 THRESHOLD OF NOTICEABLE DAMAGE 3.3.1 ANALYSIS OF FIELD EVIDENCE Building distortion related to braced excavation or tunneling is the result of both vertical and lateral ground movement. Any approach to distinguishing the limits of noticeable deformation must include treatment of both types of displacement. Unfortunately, very little information is available that relates observed building damage to measurements of lateral strain. Most construction surveys of surrounding property are performed by optical leveling. Hence, the field data available for correlation with observed damage is generally in the form of settlement measurements. Further¬ more, lateral building strains are difficult to interpret. Because of the complex nature of building deformation, the measured separation of two points on a structure frequently cannot be corrected for the tensile or compressive strains related to bending. In this section, the influence of lateral displacement will be evaluated on a comparative basis by examining the onset of noticeable damage as a function of both vertical movement alone and of combined vertical and lateral movement. Correspondingly, correlations of architectural damage with measured settlement are developed from two sources: 1) Settlement records and visual inspection associated with the underpinning of structures, and 2) Settle¬ ment records and visual inspection associated with the deformation of structures adjacent to opencutting. In the former category, observed damage is not related to lateral displacement since underpinning results almost entirely in differen¬ tial settlement. In the latter case, structural deformation results from both 3-17 vertical and lateral movements. Consequently, by comparing the two correla¬ tions, it is possible to judge how lateral displacement affects the threshold of noticeable damage. The information used to develop correlations between architectural damage and angular distortion is in the form of direct evidence, that is, all observations of damage have been documented according to both location in the structure and the maximum angular distortion that was measured. The observa¬ tions have been screened by comparing reported damage with preconstruction surveys. All the buildings included in the correlations have been inspected by the writers or their close associates. 3.3.2 ARCHITECTURAL DAMAGE RELATED TO UNDERPINNING For buildings influenced by the excavation for and installation of underpinning, evidence concerning settlement damage and angular distortion is summarized in Fig. 3.6. Angular distortions pertaining to the differential settlement measurements at 30 locations and representing 9 brick-bearing wall and 3 frame structures have been plotted. Information related to each structure is summarized in the table that accompanies the figure. Because age provides a rough measure of deterioration, especially for brick-bearing wall structures, this information is also listed. Only two instances of damage are indicated. In one case, a brick¬ bearing wall settled as the result of excavating an underpinning pit beneath it. The angular distortion, measured relative to the opposite bearing wall of the structure, was 2.4 x 10‘^ (1/417). Several cracks and a sticking door were reported. In the other case, angular distortion of 1.5 x 10’ (1/667) was measured in response to the settlement of an H-column during underpinning. \ Angular Distortion, 3-18 > «o Damaged 1/5000 1/2000 I /1000 1/600 1/200 ▲ Undomaged ■A •• •A O O • - O O -AO - A 0« • -O -O - A« -A -A A -A A O 0.0002 0.0005 0.001 0.002 0.005 Symbol Age Description • 50 To 80 Yrs. 2-4 Story Brick Bearing-Wall Structures With Timber Beams , Rubble Foundations A 35 Yrs. 13 Story Steel Frame With Concrete Floors, Infilled Plaster Partition Walls; Spread Footing Foundation 0 w 20 Yrs 6 Story Steel Frame With In-Filled Hollow Block and Plaster Walls', Spread Footing Foundotion A 45 Yrs 10 Story Steel Frame With Concrete Floors and In-Filled Brick Walls, Spreod Footing Foundation 2 Bosement Levels Fig. 3,6 Field Evidence of Architectural Damage Related to Angular Distortion for the Underpinning of Structures 3-19 Cracks in plaster partition walls between the settled and an adjacent column were noticed as high as 8 stories above the level of underpinning. On the basis of this limited data, it is difficult to set a thresh¬ old for the appearance of damage in response to underpinning settlements. The data, however, tend to support a relatively low threshold of critical distortion. In each case, the angular distortion associated with damage was smaller than the limiting value proposed by Skempton and MacDonald and Grant, et al. There are, at least, two reasons for this discrepancy. In the first place, underpinning was performed on relatively old structures that had settled in response to their own weight. The strains associated with settlement and age may have added to the strains sustained during underpinning to render an apparently low threshold for damage due to angular distortion. Secondly, the strains imposed during underpinning occur over a relatively short period of time without much benefit from creep or readjustment of building loads. When compared with correlations based on long term settlements accumulated during and after building construction, it seems reasonable that the short-term values would indicate lower limits for the onset of cracking. 3.3.3 ARCHITECTURAL DAMAGE RELATED TO MOVEMENTS CAUSED BY ADJACENT EXCAVATION For buildings adjacent to deep opencutting, evidence concerning architectural damage and angular distortion is summarized in Fig. 3.7. The maximum, angular distortions pertaining to 9 brick-bearing wall and 5 frame structures have been plotted. One of the data points, pertaining to brick¬ bearing wall structures, represents a row of 5 commercial buildings that were Angular Distortion, 3-20 > 50Yrs 9-10 Story Steel Frame With Brick Curtain Walls) Spread Footing Foundotion D 5 Yrs 10 Story Reinforced Concrete Frame ; 4 . 5 'ft. ( 1 , 4 m) Thick Reinforced Concrete Mot Foundation) 2 Bosement Levels A 10 Yrs. 9 Story Reinforced Concrete Frome ) Flat Slob Construction) Spread Footing Foundation Note Data Point Represents A Row Of 5 BuMdings Interconnected With Common Beoring Wolls Fig. 3-7 Fitld Evidence of Architectural Damage . Angular Distortion for Structures Adjacent to Braced Excavations \ 3-21 interconnected by common party walls. Information related to each structure, including its age, is listed in the table that accompanies the figure. A better understanding of the architectural damage caused by adja¬ cent excavation can be obtained by studying the detailed observations and inclinometer measurements that exist for several of the cases. Information of this nature is assembled in Figs. 3.8, 3.9 and 3.10 where displacements and observed damage are indicated with respect to scale representations of the various structures. Figure 3.8 shows a 3-story, brick-bearing wall structure that was located 50 ft 05.2 m) from the edge of a 60 ft (18.3 m) deep excavation. The soil profile at the site is composed primarily of dense sand with some interbedded stiff clay. As is common for many commercial buildings, an under¬ ground vault adjoins the structure. The vault extends 25 ft (7.6 m) from the edge of the cut. A 1/8 in. (0.32 cm) separation in the floor tiles of the display case was first noticed just after the bottom level braces had been removed from the excavation. Corresponding to this time, the surface settle¬ ments and the lateral displacement profiles at several distances from the edge of excavation are shown in the figure. The measured ground movements were taken at a location that was offset 20 ft (6.1 m) from the building front. An inspection of the building showed separations between the facade wall and all floors of the structure. These separations occurred at exactly the same loca¬ tion as the tile separation in the display case. No cracks were evident in the vault. Apparently, vertical and lateral soil movements were transmitted across the vault to the building facade. Although surface settlement at the building line was only 1/8 in. (0.32 cm), the disturbance had become large 3-22 3 Story Brick Beorinq- Wal I Structure Brick Facade Wall 1/4 in. (0.64 cm) Separation Bv»t. 2nd Floor ond Focode Wall Display Case 1/8 in.(0.32 cm) Separation in Tile 1/8 in. (0.32 cm) Separation Bwt. Bsmt. Slob and Facade Wall Scale 10 ft. Notation: • Settlement a) Detail of 3-story Building b) Lateral Displacement Profiles Fia 3.8 Architectural Damage to a 3-Story Brick- Bearing Wall Structure Settlement , ia (cm) enough to be noticed and reported by the building owner. Angular and lateral distortion along the vault were 1.0 x 10 ^ (1/1000) and 0.8 x 10 ^ (1/1250), respectively. Figure 3.9 shows the first floor level of a 6-story, steel frame structure that was located 4 ft (1.2 m) from the edge of a 55 ft (16.8 m) deep excavation. The building contains one basement level and is supported on spread footings at a depth of approximately 12 ft (3.7 m) below the ground surface. The building was not underpinned. The soil profile at the site is composed primarily of dense sand and interbedded stiff clay. Settlement of the building and lateral displacements at the edge of excavation are shown corresponding to a time just after the bottom level braces were removed from the cut. During this time, 1/32 to 3/32 in. (0.8 to 2.4 m) cracks formed along the mortar joints in a characteristic "saw-tooth" pattern. They were located in the second bay behind the front of the structure where maximum angular distortion was measured as 2.0 x lO’^ (1/500). The location and orien¬ tation of the cracks suggest that they were related to diagonal extension of the building frame. Lateral distortion of the ground surface is estimated as 1.6 X 10~^ (1/625). With time, additional cracks developed in the exterior walls between the four columns closest to the excavation. Because many of the interior building walls were covered with false paneling or display mate¬ rial , cracks were not apparent inside the structure. Figure 3.10 shows a plan view of five brick-bearing wall structures with respect to an 80-ft (24.4-m) deep excavation that was located 12 ft (3.7 m) from the building line. The structures range in size from 2 to 4 stories and contain one basement level each. The soil profile is similar to the profile Architectural Damage to a 6-Story Steel Frame Structure K£> Settlement, ia(cm) Depth, ft (m) 3-25 Fig. 3.10 Architectural Damage to a Row of Brick- Bearing Wall Structures 3-26 indicated in Fig. 3.8. All the structures were underpinned along the building line prior to excavation with jacked, pipe piles. The settlements associated with underpinning were mostly less than 3/8 in. (1.0 cm) and any related damage was confined to the area near the front of the structures. The settlements, corresponding to a time when the bottom three brace levels were removed from the cut, are shown in the form of settlement contours. In response to exca- vation the maximum angular distortion of the buildings is 1.0 x 10 (1/1000). Although the vertical movements were small, cracks and separations developed that were not noticed previously even though the buildings had been inspected before the beginning of construction and just after underpinning. Typical forms of architectural damage are indicated in the figure. Section A-A shows a 1/4 in. (0.64 cm) separation that developed in the bearing wall of the end structure at the basement level. The separation could be traced as a 1/8 to 1/4 in. (0.32 to 0.64 cm) crack in the exterior cladding of the structure at the first floor level. The separation apparently developed at the junction between two distinct sections of the wall. Section B-B shows a 1/8 in. (.0.32 cm) crack that occurred between a bearing wall and rear facade wall at the first floor level. Similar cracks were observed in three other structures. The cracks increased in width in the upper stories of the buildings. Although it is impossible to make a single, comprehensive judgment concerning these observations, several comments are offered to emphasize the salient features of the building and excavation behavior: 1. Architectural damage frequently occurred al areas of local weakness within a given structure. Ground strains, trans¬ mitted to zones of structural discontinuity, were evidenced 3-27 in cracks and separations at facade walls and between in¬ dividual sections of bearing walls. 2. At the buildings under observation, ground movements were typical of the displacements associated with adequately braced, dewatered excavations in dense sand and interbedded stiff clay. Several of the brick-bearing wall structures were underpinned prior to excavation. Total settlement at the building wall adjacent to the excavation were approximately 3/4 in. (1.9 cm), of which 40% occurred during underpinning and 60% during excavation. Lateral displacements were estimated to be in the range of 1/4 to 3/4 in. (.6 to 1.9 cm). Although the settlement of these structures was small, architectural damage was observed and reported by the occupants. By way of contrast, a 6-story, steel frame structure was not under¬ pinned even though it was located only 4 ft (1.2 m) from the edge of excavation. Although cracks were apparent in the ex¬ ternal building wall, no damage was observed during inspection of the building interior nor reported by the occupants. 3. In one instance, ground strains were transmitted across an underground vault to cause local cracks and separations at the facade wall of an adjoining structure. Although under¬ ground vaults generally are not considered part of a given structure when determining the building line, the vaults, nevertheless, can transfer ground strains to the building and cause architectural damage. 4. The largest, cumulative movements correspond with strut removal during construction of the underground structure. Correspondingly, much of the observed damage occurred during 3-28 a time when the bottom level struts were removed from the excavations. Especially when removing the lower brace levels, deep-seated ground movements are generated that can influence adjacent buildings at a significant dis¬ tance from the edge of excavation. Judging from the data summarized in Fig. 3.7 and by reviewing the previous case histories, it seems reasonable to recommend an angular distor- tion of 1.0 X 10 (1/1000) as the threshold value for architectural damage to brick-bearing wall structures adjacent to braced excavations. Only one case of damage was reported for a lower angular distortion and this corres¬ ponds to an instance when cracks were concentrated near the expansion joint of the structure. This case has been discussed in the previous section (see Fig. 3.5). Information pertaining to frame structures is limited, but it appears that an angular distortion of 1.3 x 10 (1/750) could be used as a conservative lower bound for the first appearance of damage when these buildings are adjacent to open cuts. 3.3.4 COMPARISON OF FIELD EVIDENCE Figure 3.11 compares the evidence concerning architectural damage and angular distortion as it was developed for 1) the settlement of structures under their own weight, 2) settlement caused solely by underpinning, and 3) movements associated with adjacent excavation. The summary of evidence for structures that settled under their own weight represents the combined data of Skempton and MacDonald, and Grant, et al. The limiting value of angular distortion recommended for brick-bearing wall structures adjacent to opencutting is Comparison of Field Evidence Related to Architectural Damage Angulor Dittortion , dv CO CO i ip iXi -“ff (s ' 3 I|°S Q 3 ) > L o ® " - 3 ) - Q O 5 5 •