Harbour Cay Condominium (March 27, 1981) Jason D. Kukorlo, BAE/MAE, Penn State, 2009
Abstract
The collapse of Harbour Cay Condominium, a five-story flat-plate reinforced concrete residential building in Cocoa Beach, Florida, occurred around 3:00 PM on March 27, 1981, due to a punching shear failure. Workers were placing the concrete for the roof slab when failure of the slab at one column initiated a collapse of the entire 5th floor. The 5th floor slab fell onto the floors below it, which led to the progressive collapse of the entire structure. Eleven workers were killed and 23 were injured (Lew et al. 1982, p. vii-ix).
Investigation teams concluded that both errors in design and construction led to the tragic failure. On the design side, it was found that the slab thickness fell short of ACI code requirements. The slab was 8 inches thick whereas the minimum slab thickness required by code to account for punching shear was 11 inches. Punching shear was simply not considered in the design (Lew et al. 1982, p. ix). The structural engineer was a retired NASA engineer who neglected to perform several other necessary concrete calculations as well, including checks for deflection, beam shear, and column reinforcement spacing (Delatte 2009, p. 153). Steel reinforcement in the columns was over-congested, which made it difficult for the concrete to flow around the rebar (Feld 1997, p. 273). On the construction side, the chairs used to support the top reinforcement in the slab were too short, which reduced the effective slab depth and hence decreased the slab punching shear strength (Lew et al. 1982, p. 65). Studies found that both the design and construction errors contributed fairly equally to the failure and that the collapse probably would not have occurred had only one of these errors been made (Delatte 2009, p. 153).
Severe cracking of the floors and excessive deflections were brought to the attention of the structural engineer prior to the collapse, but he still confirmed that the structure had sufficient strength (Kaminetzky 1991, p. 74). Construction should have stopped immediately after the warning signs of “spider-web-type cracks” and deflections up to 1 ¾”. Improper reshoring procedures may have also contributed to the structural failure. Additionally, testing showed that the shear strength of the slab near columns was exceeded at numerous locations, and some of the concrete used had a nonuniform consistency (Lew et al. 1982, p. iii-vii). The analysis of the structure concluded that a punching shear failure occurred on the 5th floor at one of the interior columns, which led to a progressive collapse of the entire slab and hence the entire structure (Lew et al. 1982, p. ix).
Figure 1: Harbour Cay Condominium Before the Collapse (Image courtesy of National Bureau of Standards)
Key Words
Punching shear, flat plate, progressive collapse, concrete strength, failure during construction
Brief Summary of the Events Leading Up to the Collapse
Harbour Cay Condominium was a relatively low-rise building with a length of 242 feet and width of 58 feet (Kaminetzky 1991, p. 72). Figure 1 shows an aerial view of the building before the collapse occurred. When completed, the condominium complex was to feature two five-story buildings, each with forty-five units and twenty-eight townhouses (Wright 2006, p. 113). Flying forms made up of preassembled plywood decks on aluminum trusses were used to construct the 8-inch-thick floor slabs (Lew et al. 1982, p. 1). The maximum span of the slabs was 27’-8” and 22’-2” in the two directions. A structurally detached elevator tower was located at the east end of the building, and stairwells were located at the north and south ends. Typical exterior columns were 10 in. x 12 in., and typical interior columns were 10 in. x 18 in (Delatte 2009, p. 152). The columns along the exterior ends of the stairwells were 8 in. x 12 in. Story heights were 8’-8” from top of slab to top of slab (Lew et al. 1982, p. 5). All above-grade concrete was specified to have a compressive strength of 4,000 psi (Delatte 2009, p. 152). Pile caps with two to nine piles each supported the columns, and continuous wall footings connected the exterior pile caps (Lew et al. 1982, p. 5). While the roof was planned to be cast in one day, floor slabs were cast in two parts with each comprising one-half of the area of the entire floor (Lew et al. 1982, p. 1). Floors were cast at a rate of approximately one floor per week (Lew et al. 1982, p. 11).
The structure was almost complete at the time of the collapse. Concrete had already been placed on seven of the nine total flying forms for the roof (Lew et al. 1982, p. vii). Workers were finishing the roof slab when they heard a loud crack that sounded like wood splitting. Some eyewitnesses claimed that the center of the fifth floor came down first, while others reported the fourth floor as falling first with the fifth floor collapsing afterward (Kamitnetzky 1991, p. 72). The top floors then fell onto the floors below, causing a progressive collapse of the entire structure.
Days prior to the collapse, workers observed and reported several spider-web-type cracks in the floor slabs and excessive deflections of up to 1 ¾”. The structural engineer was informed of the problems and was asked to recheck his design, which he did. However, he still ensured that his design was adequate (Kaminetzky 1991, p. 74). Work on the project proceeded despite these warning signs of a potential failure.
Figure 2: Location of Workers in Building at Time of the Collapse (Image courtesy of National Bureau of Standards)
At the time of the collapse, the exterior walls for the first two floors at the south end of the building were in place, and masonry units were piled up on the third floor. Masonry walls for the first three floors in the north end of the building had been constructed, and masonry units were piled up on the fourth floor (Lew et al. 1982, p. 11). When the structure collapsed, reshores were located in the second, third, and fourth stories, but the exact number of reshores present was questionable (Lew et al. 1982, p. 12). The only reshores on the first floor were a few peripheral reshores under the walkways and balconies (Lew et al. 1982, p. 9).
There were 36 workers located throughout the building when the structure collapsed. Figure 2 shows the location of the workers in the building at the time of collapse. Two workers on the ground floor were installing window unit framing, while most of the workers on the third floor were completing masonry walls. One worker in the fourth story was cleaning up debris and two others were making leveling adjustments in the formwork. Workers on the roof were finishing the placement of the concrete slab and were located near the north two bays which had not been poured yet. Workers stated that at the time of the collapse there was no concrete being delivered to the roof (Lew et al. 1982, p. 11).
Causes of the Failure
Figure 3: Harbour Cay Condominium Collapse (Image courtesy of National Bureau of Standards)
Design Errors
The collapse of Harbour Cay Condominium was attributed to both design errors and construction errors. On the design side, the 8-inch slab thickness fell short of the minimum slab thickness of 11 in. required by ACI code to resist punching shear for the given loads, spans, and column sizes (Kaminetzky 1991, p. 75). Punching shear is the most common mode of failure for flat-plate structures, yet punching shear calculations were simply omitted from the design (Feld 1997, p. 274). A calculation for minimum slab thickness was also not performed. After the collapse, many of the columns remained standing with the floor slabs stacked on top of each other on the ground. This showed further strong evidence of a punching shear failure (Kaminetzky 1991, p. 75). Images of Harbour Cay Condominium after the collapse can be seen in Figure 3 (right) and Figure 5 (below). The only significant loads on the structure at the time of collapse were gravity loads (Feld 1997, p. 274). No evidence of overturning or sidesway of the building was found (Lew et al. 1982, p. 36).
“A punching shear failure happens when the concrete floor slab cracks and breaks away from its column connection. It’s as though you are poking the column through the floor slab like a pencil through a piece of paper” (Modern Marvels 2004). The punching shear strength of a flat slab for a simplified case of an interior column is:
Vc = 4((f’c)^(1/2))(b0)(d)
f’c = 28-day cylinder compressive strength of the concrete
d = depth of slab (measured from the bottom of the slab to the reinforcing
steel location)
b0 = the perimeter of the failure surface around the column measured at
distance “d” from the face of the column
(Delatte 2009, p. 140)
To increase punching shear strength, the shear perimeter should be increased by using larger columns or column capitals. Also, some of the top (negative-moment tension) steel of slabs should pass through columns in all punching shear cases (Nawy 2008, p. 8-27).
The structural engineer was a retired NASA engineer who hired another retired NASA engineer to perform the calculations (Feld 1997, p. 274). Delatte made a point that “…structural engineering isn’t rocket science. Evidently, it is considerably more difficult” (Delatte 2009, p. 153).
Overall, design errors included:
-There were no calculations for deflection or minimum thickness provisions.
-There were no calculations for punching shear or beam shear.
-There were no code checks for column reinforcement spacing.
-Calculations used Grade 40 steel whereas the structural drawings specified Grade 60 steel.
-There were no actual calculations for the effective depth of slab flexural reinforcement. A constant multiplier of computed moments was used instead (Lew et al. 1982, p. 6).
-Congested column reinforcement prevented concrete from flowing around the steel bars and thus caused a deficient bond between the reinforcement and concrete (Feld 1997, p. 273).
Figure 4: Configuration of Chairs and Top Reinforcement in Column Strip (Image courtesy of National Bureau of Standards)
Construction Errors
On the construction side, the top reinforcement steel was placed too low, which reduced the effective slab depth and hence the punching shear capacity of the slab. Figure 4 shows the configuration of chairs and top reinforcement in a column strip within the building. The top reinforcement bars were placed on chairs that were only 4 ½” high, which reduced the effective slab depth “d” from 6.3 in. to 5.3 in. Hence, the top cover was increased to 1 5/8” whereas it was designed to be ¾” (Delatte 2009, p. 153). The NBS investigation also found that bottom slab bars were not placed through many columns and that the slabs broke away from the columns where the slabs and columns met. In addition, some vertical reinforcement was found to have been severely bent during fabrication (Lew et al. 1982, p. 32). Laboratory-cured test cylinders were used instead of field-cured test cylinders to determine the actual strength of slabs prior to the stripping of formwork (Kaminetzky 1991, p. 76).
The NBS report also included interviews with workers and witnesses who were present at the time of the collapse. Many disagreements exist as to the location and amount of reshores present in the structure when it collapsed. Some workers said they saw bowed reshores, and others even said they saw reshores break when concrete was being placed on the roof. Due to the numerous discrepancies in the workers’ accounts, it is basically impossible to determine the exact layout of reshores at the time of collapse (Lew et al. 1982, p. 33). Many workers stated that the spider-web-type cracks were noticed once the flying forms were removed. Most cracks were located near midspans and around columns, and some were said to have extended 4 to 5 inches into the floor slabs. Excessive deflections were reported once the forms were removed. A 1 ¾” deflection was noted in the end apartment on the north side of the building on the second floor (Lew et al. 1982, p. 34). Workers also noted that some of the concrete from the on-site batch plant had a non-uniform consistency and was difficult to finish (Lew et al. 1982, p. 37). One worker stated, “Twenty-two years I’ve been pouring concrete and they’ve never pulled the forms in two days like they did here. They usually set there for a week or 10 days” (Montgomery 1981).
Some investigators wondered why the structure had not collapsed earlier. The shoring and reshoring methods used provided the answer. Shores and reshores initially supported the dead loads of the structure and transferred the loads to the ground. Once the reshores below the first floor level were removed, the concrete slabs were forced to carry the weight of the structure through their punching shear capacity at the columns (Kaminetzky 1991, p. 75).
Conflicting Accounts
There were not really any major conflicting accounts of the failure or dissenting opinions of the cause of the failure. Most investigators agree that the main cause of the Harbour Cay Condominium collapse was a punching shear failure which led to a progressive collapse of the entire structure. Since many of the columns were still standing after the collapse, it was a clear indication that the structure collapsed due to a punching shear failure. Investigators agree that punching shear was not accounted for in the design of the structure, and that reinforcing bars were placed too low in the slab. One small argument relates to the concept of design error versus construction error. The National Bureau of Standards concluded that both the design error of the 8-inch slab and the construction error of the inadequate effective depth of reinforcing steel contributed fairly equally to the collapse of the structure. The NBS determined that the structure would have probably not collapsed had only one of these errors been made (Delatte 2009, p. 153). Feld and Carper, however, state “Some construction deficiencies were noted, but the design error related to punching shear alone was clearly sufficient to bring about the collapse” (Feld 1997, p. 274). Basically, Feld and Carper claim that the punching shear design error would have caused the collapse itself, whereas the NBS stated that the building would have stayed standing had only the design error occurred. Also, as previously discussed, eyewitness accounts also offer several dissenting opinions as to the location and number of reshores in the building at the time of collapse and whether the fourth floor or fifth floor came down first.
Prevention of the Failure
It is evident that the failure could have easily been prevented. Had the simple routine concrete design checks for punching shear and minimum slab thickness been made, the punching shear failure could have been avoided. The most economical way to increase the punching shear capacity of the slabs would have been to increase the size of the columns. This would also have created more space for casting concrete between the vertical column reinforcement bars (Kaminetzky 1991, p. 75). Increasing the thickness of the slab would have required much more concrete than increasing the size of the columns. Hence, increasing the column sizes would have provided a more economical solution (Delatte 2009, p. 155). Delatte also mentions how the Harbour Cay Condominium failure could have been prevented had the engineers learned the lessons of the 2000 Commonwealth Avenue collapse, which was also caused by a punching shear failure (Delatte 2009, p. 154).
In addition, paying attention to warning signs of a potential collapse is critical. All work on the building should have stopped after the excessive deflections and spider-web-type cracks had formed. Instead, work on the building continued without properly addressing these obvious signs of possible failure.
As mentioned above, the NBS investigators found that if only the main design error or the main construction error had been made, the structure would probably not have collapsed. Therefore, had one of the major errors involved with the structure been eliminated, the building may possibly still be standing.
Lessons Learned
Several lessons can be learned from the collapse of Harbour Cay Condominium. First, punching shear strength must be checked when designing flat slabs, for punching shear is the most common mode of failure for concrete flat slabs. Second, minimum depth of a flat slab much be checked to account for deflection and strength requirements. Next, it is crucial to place reinforcing bars directly within the column periphery to help prevent progressive collapse. This can be done at no additional cost. Furthermore, proper design of formwork, shoring and reshoring plans and schedules, and procedures to verify minimum stripping strength of the concrete by professionals are essential for successful field construction control. Another important lesson is that all work on a project must be stopped if warning signs of potential failure are encountered. Workers should evacuate the building immediately, and professional evaluation of the problems must be performed before work can be resumed. Finally, it is important to use proper test methods to determine the in-place strength of concrete in cold weather. Field-cured test cylinders should be used instead of laboratory-cured test cylinders. The level of construction carelessness also increases during the winter (Kaminetzky 1991, pp. 77-78).
The building industry can also learn the consequences of a catastrophic failure like the Harbour Cay Condominium collapse. The primary structural engineer on the project, Harold Meeler, surrendered his license and said he would never practice again. Meeler said he would pay the maximum fine of $3,000 to avoid a hearing on the collapse of the structure (Engineer 1981). The other structural engineer also surrendered his license and will never practice in the state of Florida again. The Florida Department of Professional Regulation charged five of the parties involved in the project with negligence. Additionally, two contractors were disciplined, and the architect was suspended from practicing in Florida for ten years. We must remember that major failures in low-rise projects are still possible despite all of the knowledge available to avoid them (Feld 1997, p. 274).
Changes Made in the Industry
The state of Florida strengthened its safety laws after the Harbour Cay Condominium collapse, requiring more on-site inspections by engineers and more scrutiny of construction plans (Modern Marvels 2004). The failure also raised awareness that punching shear failures are the most common type of failure of concrete flat slabs, and accounting for punching shear during the design stage is crucial. The collapse also demonstrated that major catastrophes can still occur with low-rise buildings and not just high-rise structures.
Figure 5: Harbour Cay Condominium Collapse (Image courtesy of National Bureau of Standards)
Could This Collapse Occur Again?
It seems that making a major design error such as omitting punching shear calculations for a concrete flat slab would be rather unlikely today. However, punching shear failures can still occur, especially if formwork is removed before the concrete gains full strength or improper shoring and reshoring is used during construction. The 2000 Commonwealth Avenue collapse and the Bailey’s Crossroads collapse were similar case studies that were both caused by punching shear failures (Delatte 2009, p. 144). Fourteen years after the Harbour Cay catastrophe, the Sampoong Department Store in Seoul, South Korea, collapsed due to a punching shear failure, killing 502 people and injuring more than 900 others. Like the Harbour Cay structure, the store had flat slabs and reinforced concrete columns. Another floor was added to the four-story building in 1989, and the structure collapsed in 1995 after ten tons of air conditioning units were added to the roof. Like the Harbour Cay building, cracks in the concrete were brought to the Sampoong Store owner’s attention days before the collapse, but the warning signs were ignored (Modern Marvels 2004). All of these case studies demonstrate the importance of punching shear capacity and how a punching shear failure can cause a progressive collapse. Adequate slab thickness, concrete strength, and proper placement of reinforcement are essential to prevent such a failure (Delatte 2009, p. 144). The importance of not ignoring warning signs of a potential disaster is evident as well.
Conclusion
The Harbour Cay Condominium collapse demonstrates the consequences of improper design and construction procedures. A punching shear failure on the fifth floor initiated a progressive collapse of the entire structure. Punching shear calculations were omitted by the structural engineer when the structure was designed. Reinforcement bars were placed too low in the concrete slabs, which reduced the effective depth of the slabs and hence reduced the overall strength of the slabs as well. Warning signs of a potential failure were brought to the attention of supervisors and the structural engineer but were basically ignored. The Harbour Cay disaster could have easily been prevented had simple design checks and careful construction techniques been performed.
Bibliography
1. Delatte, Norbert J. Beyond Failure: Forensic Case Studies for Civil Engineers.
ASCE Press, 2009, p. 149-155.
2. “Engineer in Building Collapse Gives Up His Florida License.” New York Times,
September 13, 1981.
3. Feld, Jacob and Kenneth Carper, K. Construction Failure. 2nd Ed., John Wiley &
Sons, New York, N. Y., 1997, p. 271-274.
4. Kaminetzky, D. Design and Construction Failures: Lessons from Forensic
Investigations. McGraw-Hill, New York, N. Y., 1991, p. 72-78.
5. Lew, H. S. et al. “Investigation of Construction Failure of Harbour Cay
Condominium in Cocoa Beach, Florida.” Rep., U.S. Dept. of Comm., Nat. Bureau
of Standards, S/N 003-003-02405-8, Washington, D. C., 1982.
6. Modern Marvels Engineering Disasters 15. History Channel, A & E Home Video,
2004.
7. Montgomery, Paul L. “Two Still Missing in Condominium Collapse.” New York
Times, March 29, 1981.
8. Nawy, Edward G. Concrete Construction Engineering Handbook. CRC Press, Roca
Raton, FL, 2008, p. 8-27, 8-37 to 8-40.
Table of Contents
Jason D. Kukorlo, BAE/MAE, Penn State, 2009
Abstract
The collapse of Harbour Cay Condominium, a five-story flat-plate reinforced concrete residential building in Cocoa Beach, Florida, occurred around 3:00 PM on March 27, 1981, due to a punching shear failure. Workers were placing the concrete for the roof slab when failure of the slab at one column initiated a collapse of the entire 5th floor. The 5th floor slab fell onto the floors below it, which led to the progressive collapse of the entire structure. Eleven workers were killed and 23 were injured (Lew et al. 1982, p. vii-ix).
Investigation teams concluded that both errors in design and construction led to the tragic failure. On the design side, it was found that the slab thickness fell short of ACI code requirements. The slab was 8 inches thick whereas the minimum slab thickness required by code to account for punching shear was 11 inches. Punching shear was simply not considered in the design (Lew et al. 1982, p. ix). The structural engineer was a retired NASA engineer who neglected to perform several other necessary concrete calculations as well, including checks for deflection, beam shear, and column reinforcement spacing (Delatte 2009, p. 153). Steel reinforcement in the columns was over-congested, which made it difficult for the concrete to flow around the rebar (Feld 1997, p. 273). On the construction side, the chairs used to support the top reinforcement in the slab were too short, which reduced the effective slab depth and hence decreased the slab punching shear strength (Lew et al. 1982, p. 65). Studies found that both the design and construction errors contributed fairly equally to the failure and that the collapse probably would not have occurred had only one of these errors been made (Delatte 2009, p. 153).
Severe cracking of the floors and excessive deflections were brought to the attention of the structural engineer prior to the collapse, but he still confirmed that the structure had sufficient strength (Kaminetzky 1991, p. 74). Construction should have stopped immediately after the warning signs of “spider-web-type cracks” and deflections up to 1 ¾”. Improper reshoring procedures may have also contributed to the structural failure. Additionally, testing showed that the shear strength of the slab near columns was exceeded at numerous locations, and some of the concrete used had a nonuniform consistency (Lew et al. 1982, p. iii-vii). The analysis of the structure concluded that a punching shear failure occurred on the 5th floor at one of the interior columns, which led to a progressive collapse of the entire slab and hence the entire structure (Lew et al. 1982, p. ix).
Key Words
Punching shear, flat plate, progressive collapse, concrete strength, failure during construction
Brief Summary of the Events Leading Up to the Collapse
Harbour Cay Condominium was a relatively low-rise building with a length of 242 feet and width of 58 feet (Kaminetzky 1991, p. 72). Figure 1 shows an aerial view of the building before the collapse occurred. When completed, the condominium complex was to feature two five-story buildings, each with forty-five units and twenty-eight townhouses (Wright 2006, p. 113). Flying forms made up of preassembled plywood decks on aluminum trusses were used to construct the 8-inch-thick floor slabs (Lew et al. 1982, p. 1). The maximum span of the slabs was 27’-8” and 22’-2” in the two directions. A structurally detached elevator tower was located at the east end of the building, and stairwells were located at the north and south ends. Typical exterior columns were 10 in. x 12 in., and typical interior columns were 10 in. x 18 in (Delatte 2009, p. 152). The columns along the exterior ends of the stairwells were 8 in. x 12 in. Story heights were 8’-8” from top of slab to top of slab (Lew et al. 1982, p. 5). All above-grade concrete was specified to have a compressive strength of 4,000 psi (Delatte 2009, p. 152). Pile caps with two to nine piles each supported the columns, and continuous wall footings connected the exterior pile caps (Lew et al. 1982, p. 5). While the roof was planned to be cast in one day, floor slabs were cast in two parts with each comprising one-half of the area of the entire floor (Lew et al. 1982, p. 1). Floors were cast at a rate of approximately one floor per week (Lew et al. 1982, p. 11).
The structure was almost complete at the time of the collapse. Concrete had already been placed on seven of the nine total flying forms for the roof (Lew et al. 1982, p. vii). Workers were finishing the roof slab when they heard a loud crack that sounded like wood splitting. Some eyewitnesses claimed that the center of the fifth floor came down first, while others reported the fourth floor as falling first with the fifth floor collapsing afterward (Kamitnetzky 1991, p. 72). The top floors then fell onto the floors below, causing a progressive collapse of the entire structure.
Days prior to the collapse, workers observed and reported several spider-web-type cracks in the floor slabs and excessive deflections of up to 1 ¾”. The structural engineer was informed of the problems and was asked to recheck his design, which he did. However, he still ensured that his design was adequate (Kaminetzky 1991, p. 74). Work on the project proceeded despite these warning signs of a potential failure.
At the time of the collapse, the exterior walls for the first two floors at the south end of the building were in place, and masonry units were piled up on the third floor. Masonry walls for the first three floors in the north end of the building had been constructed, and masonry units were piled up on the fourth floor (Lew et al. 1982, p. 11). When the structure collapsed, reshores were located in the second, third, and fourth stories, but the exact number of reshores present was questionable (Lew et al. 1982, p. 12). The only reshores on the first floor were a few peripheral reshores under the walkways and balconies (Lew et al. 1982, p. 9).
There were 36 workers located throughout the building when the structure collapsed. Figure 2 shows the location of the workers in the building at the time of collapse. Two workers on the ground floor were installing window unit framing, while most of the workers on the third floor were completing masonry walls. One worker in the fourth story was cleaning up debris and two others were making leveling adjustments in the formwork. Workers on the roof were finishing the placement of the concrete slab and were located near the north two bays which had not been poured yet. Workers stated that at the time of the collapse there was no concrete being delivered to the roof (Lew et al. 1982, p. 11).
Causes of the Failure
Design Errors
The collapse of Harbour Cay Condominium was attributed to both design errors and construction errors. On the design side, the 8-inch slab thickness fell short of the minimum slab thickness of 11 in. required by ACI code to resist punching shear for the given loads, spans, and column sizes (Kaminetzky 1991, p. 75). Punching shear is the most common mode of failure for flat-plate structures, yet punching shear calculations were simply omitted from the design (Feld 1997, p. 274). A calculation for minimum slab thickness was also not performed. After the collapse, many of the columns remained standing with the floor slabs stacked on top of each other on the ground. This showed further strong evidence of a punching shear failure (Kaminetzky 1991, p. 75). Images of Harbour Cay Condominium after the collapse can be seen in Figure 3 (right) and Figure 5 (below). The only significant loads on the structure at the time of collapse were gravity loads (Feld 1997, p. 274). No evidence of overturning or sidesway of the building was found (Lew et al. 1982, p. 36).
“A punching shear failure happens when the concrete floor slab cracks and breaks away from its column connection. It’s as though you are poking the column through the floor slab like a pencil through a piece of paper” (Modern Marvels 2004). The punching shear strength of a flat slab for a simplified case of an interior column is:
Vc = 4((f’c)^(1/2))(b0)(d)
f’c = 28-day cylinder compressive strength of the concrete
d = depth of slab (measured from the bottom of the slab to the reinforcing
steel location)
b0 = the perimeter of the failure surface around the column measured at
distance “d” from the face of the column
(Delatte 2009, p. 140)
To increase punching shear strength, the shear perimeter should be increased by using larger columns or column capitals. Also, some of the top (negative-moment tension) steel of slabs should pass through columns in all punching shear cases (Nawy 2008, p. 8-27).
The structural engineer was a retired NASA engineer who hired another retired NASA engineer to perform the calculations (Feld 1997, p. 274). Delatte made a point that “…structural engineering isn’t rocket science. Evidently, it is considerably more difficult” (Delatte 2009, p. 153).
Overall, design errors included:
-There were no calculations for deflection or minimum thickness provisions.
-There were no calculations for punching shear or beam shear.
-There were no code checks for column reinforcement spacing.
-Calculations used Grade 40 steel whereas the structural drawings specified Grade 60 steel.
-There were no actual calculations for the effective depth of slab flexural reinforcement. A constant multiplier of computed moments was used instead (Lew et al. 1982, p. 6).
-Congested column reinforcement prevented concrete from flowing around the steel bars and thus caused a deficient bond between the reinforcement and concrete (Feld 1997, p. 273).
Construction Errors
On the construction side, the top reinforcement steel was placed too low, which reduced the effective slab depth and hence the punching shear capacity of the slab. Figure 4 shows the configuration of chairs and top reinforcement in a column strip within the building. The top reinforcement bars were placed on chairs that were only 4 ½” high, which reduced the effective slab depth “d” from 6.3 in. to 5.3 in. Hence, the top cover was increased to 1 5/8” whereas it was designed to be ¾” (Delatte 2009, p. 153). The NBS investigation also found that bottom slab bars were not placed through many columns and that the slabs broke away from the columns where the slabs and columns met. In addition, some vertical reinforcement was found to have been severely bent during fabrication (Lew et al. 1982, p. 32). Laboratory-cured test cylinders were used instead of field-cured test cylinders to determine the actual strength of slabs prior to the stripping of formwork (Kaminetzky 1991, p. 76).
The NBS report also included interviews with workers and witnesses who were present at the time of the collapse. Many disagreements exist as to the location and amount of reshores present in the structure when it collapsed. Some workers said they saw bowed reshores, and others even said they saw reshores break when concrete was being placed on the roof. Due to the numerous discrepancies in the workers’ accounts, it is basically impossible to determine the exact layout of reshores at the time of collapse (Lew et al. 1982, p. 33). Many workers stated that the spider-web-type cracks were noticed once the flying forms were removed. Most cracks were located near midspans and around columns, and some were said to have extended 4 to 5 inches into the floor slabs. Excessive deflections were reported once the forms were removed. A 1 ¾” deflection was noted in the end apartment on the north side of the building on the second floor (Lew et al. 1982, p. 34). Workers also noted that some of the concrete from the on-site batch plant had a non-uniform consistency and was difficult to finish (Lew et al. 1982, p. 37). One worker stated, “Twenty-two years I’ve been pouring concrete and they’ve never pulled the forms in two days like they did here. They usually set there for a week or 10 days” (Montgomery 1981).
Some investigators wondered why the structure had not collapsed earlier. The shoring and reshoring methods used provided the answer. Shores and reshores initially supported the dead loads of the structure and transferred the loads to the ground. Once the reshores below the first floor level were removed, the concrete slabs were forced to carry the weight of the structure through their punching shear capacity at the columns (Kaminetzky 1991, p. 75).
Conflicting Accounts
There were not really any major conflicting accounts of the failure or dissenting opinions of the cause of the failure. Most investigators agree that the main cause of the Harbour Cay Condominium collapse was a punching shear failure which led to a progressive collapse of the entire structure. Since many of the columns were still standing after the collapse, it was a clear indication that the structure collapsed due to a punching shear failure. Investigators agree that punching shear was not accounted for in the design of the structure, and that reinforcing bars were placed too low in the slab. One small argument relates to the concept of design error versus construction error. The National Bureau of Standards concluded that both the design error of the 8-inch slab and the construction error of the inadequate effective depth of reinforcing steel contributed fairly equally to the collapse of the structure. The NBS determined that the structure would have probably not collapsed had only one of these errors been made (Delatte 2009, p. 153). Feld and Carper, however, state “Some construction deficiencies were noted, but the design error related to punching shear alone was clearly sufficient to bring about the collapse” (Feld 1997, p. 274). Basically, Feld and Carper claim that the punching shear design error would have caused the collapse itself, whereas the NBS stated that the building would have stayed standing had only the design error occurred. Also, as previously discussed, eyewitness accounts also offer several dissenting opinions as to the location and number of reshores in the building at the time of collapse and whether the fourth floor or fifth floor came down first.
Prevention of the Failure
It is evident that the failure could have easily been prevented. Had the simple routine concrete design checks for punching shear and minimum slab thickness been made, the punching shear failure could have been avoided. The most economical way to increase the punching shear capacity of the slabs would have been to increase the size of the columns. This would also have created more space for casting concrete between the vertical column reinforcement bars (Kaminetzky 1991, p. 75). Increasing the thickness of the slab would have required much more concrete than increasing the size of the columns. Hence, increasing the column sizes would have provided a more economical solution (Delatte 2009, p. 155). Delatte also mentions how the Harbour Cay Condominium failure could have been prevented had the engineers learned the lessons of the 2000 Commonwealth Avenue collapse, which was also caused by a punching shear failure (Delatte 2009, p. 154).
In addition, paying attention to warning signs of a potential collapse is critical. All work on the building should have stopped after the excessive deflections and spider-web-type cracks had formed. Instead, work on the building continued without properly addressing these obvious signs of possible failure.
As mentioned above, the NBS investigators found that if only the main design error or the main construction error had been made, the structure would probably not have collapsed. Therefore, had one of the major errors involved with the structure been eliminated, the building may possibly still be standing.
Lessons Learned
Several lessons can be learned from the collapse of Harbour Cay Condominium. First, punching shear strength must be checked when designing flat slabs, for punching shear is the most common mode of failure for concrete flat slabs. Second, minimum depth of a flat slab much be checked to account for deflection and strength requirements. Next, it is crucial to place reinforcing bars directly within the column periphery to help prevent progressive collapse. This can be done at no additional cost. Furthermore, proper design of formwork, shoring and reshoring plans and schedules, and procedures to verify minimum stripping strength of the concrete by professionals are essential for successful field construction control. Another important lesson is that all work on a project must be stopped if warning signs of potential failure are encountered. Workers should evacuate the building immediately, and professional evaluation of the problems must be performed before work can be resumed. Finally, it is important to use proper test methods to determine the in-place strength of concrete in cold weather. Field-cured test cylinders should be used instead of laboratory-cured test cylinders. The level of construction carelessness also increases during the winter (Kaminetzky 1991, pp. 77-78).
The building industry can also learn the consequences of a catastrophic failure like the Harbour Cay Condominium collapse. The primary structural engineer on the project, Harold Meeler, surrendered his license and said he would never practice again. Meeler said he would pay the maximum fine of $3,000 to avoid a hearing on the collapse of the structure (Engineer 1981). The other structural engineer also surrendered his license and will never practice in the state of Florida again. The Florida Department of Professional Regulation charged five of the parties involved in the project with negligence. Additionally, two contractors were disciplined, and the architect was suspended from practicing in Florida for ten years. We must remember that major failures in low-rise projects are still possible despite all of the knowledge available to avoid them (Feld 1997, p. 274).
Changes Made in the Industry
The state of Florida strengthened its safety laws after the Harbour Cay Condominium collapse, requiring more on-site inspections by engineers and more scrutiny of construction plans (Modern Marvels 2004). The failure also raised awareness that punching shear failures are the most common type of failure of concrete flat slabs, and accounting for punching shear during the design stage is crucial. The collapse also demonstrated that major catastrophes can still occur with low-rise buildings and not just high-rise structures.
Could This Collapse Occur Again?
It seems that making a major design error such as omitting punching shear calculations for a concrete flat slab would be rather unlikely today. However, punching shear failures can still occur, especially if formwork is removed before the concrete gains full strength or improper shoring and reshoring is used during construction. The 2000 Commonwealth Avenue collapse and the Bailey’s Crossroads collapse were similar case studies that were both caused by punching shear failures (Delatte 2009, p. 144). Fourteen years after the Harbour Cay catastrophe, the Sampoong Department Store in Seoul, South Korea, collapsed due to a punching shear failure, killing 502 people and injuring more than 900 others. Like the Harbour Cay structure, the store had flat slabs and reinforced concrete columns. Another floor was added to the four-story building in 1989, and the structure collapsed in 1995 after ten tons of air conditioning units were added to the roof. Like the Harbour Cay building, cracks in the concrete were brought to the Sampoong Store owner’s attention days before the collapse, but the warning signs were ignored (Modern Marvels 2004). All of these case studies demonstrate the importance of punching shear capacity and how a punching shear failure can cause a progressive collapse. Adequate slab thickness, concrete strength, and proper placement of reinforcement are essential to prevent such a failure (Delatte 2009, p. 144). The importance of not ignoring warning signs of a potential disaster is evident as well.Conclusion
The Harbour Cay Condominium collapse demonstrates the consequences of improper design and construction procedures. A punching shear failure on the fifth floor initiated a progressive collapse of the entire structure. Punching shear calculations were omitted by the structural engineer when the structure was designed. Reinforcement bars were placed too low in the concrete slabs, which reduced the effective depth of the slabs and hence reduced the overall strength of the slabs as well. Warning signs of a potential failure were brought to the attention of supervisors and the structural engineer but were basically ignored. The Harbour Cay disaster could have easily been prevented had simple design checks and careful construction techniques been performed.
Bibliography
1. Delatte, Norbert J. Beyond Failure: Forensic Case Studies for Civil Engineers.
ASCE Press, 2009, p. 149-155.
2. “Engineer in Building Collapse Gives Up His Florida License.” New York Times,
September 13, 1981.
3. Feld, Jacob and Kenneth Carper, K. Construction Failure. 2nd Ed., John Wiley &
Sons, New York, N. Y., 1997, p. 271-274.
4. Kaminetzky, D. Design and Construction Failures: Lessons from Forensic
Investigations. McGraw-Hill, New York, N. Y., 1991, p. 72-78.
5. Lew, H. S. et al. “Investigation of Construction Failure of Harbour Cay
Condominium in Cocoa Beach, Florida.” Rep., U.S. Dept. of Comm., Nat. Bureau
of Standards, S/N 003-003-02405-8, Washington, D. C., 1982.
6. Modern Marvels Engineering Disasters 15. History Channel, A & E Home Video,
2004.
7. Montgomery, Paul L. “Two Still Missing in Condominium Collapse.” New York
Times, March 29, 1981.
8. Nawy, Edward G. Concrete Construction Engineering Handbook. CRC Press, Roca
Raton, FL, 2008, p. 8-27, 8-37 to 8-40.
9. Wright, E. Lynne. Disasters and Heroic Rescues of Florida: True Stories of Tragedy and Survival. Morris Book Publishing, LLC., 2006, p. 113-122.