Skip to main content

Full text of "DTIC ADA560616: Performance of Blast-Damaged Steel Connections in Progressive Collapse"

See other formats


PERFORMANCE OF BLAST-DAMAGED STEEL CONNECTIONS 

IN PROGRESSIVE COLLAPSE 


John M. H. Puryear, Protection Engineering Consultants, Austin, TX 
Kirk A. Marchand, Protection Engineering Consultants, Austin, TX 
David J. Stevens, Protection Engineering Consultants, San Antonio, TX 
Doug A. Sunshine, Defense Threat Reduction Agency, Washington, DC 

Abstract: In progressive collapse events initiated by explosive threats, damage to some 
connections from blast loading of the structure is likely. In this paper, the results from 
ten tests of blast-damaged and undamaged beam-to-column connections are examined. 
An evaluation procedure leading to comparisons of performance based on rotational ca¬ 
pacity of both blast damaged and undamaged capacity is presented. 

INTRODUCTION 

Connections in steel buildings provide much of the structural continuity, redundancy, and 
ductility required to resist progressive collapse. Premature connection failure causes 
structural discontinuities and reduces resistance provided by redundant load paths. Fur¬ 
thermore, brittle connections prevent structural members from developing their full duc¬ 
tility, which is critical to arresting the accelerated mass of a building in a progressive col¬ 
lapse event. Therefore, an accurate understanding of connection performance is essential 
to structural analysis for progressive collapse. 

To better understand connection performance in progressive collapse, a series of push¬ 
down tests of blast-damaged and undamaged beam-to-column connections were per¬ 
formed (1). Ten connections representing five connection types were tested. These tests 
were part of a multi-phase program including blast tests of steel columns and base plates 
( 2 , 3 ). 

In this paper, the multi-phase test program is briefly discussed, with an emphasis on the 
beam-to-column connection tests. The evaluation procedure used to analyze co nn ection 
performance is then illustrated for one of the ten connections. Connection performance is 
compared by type based on rotational capacity. Finally, conclusions are drawn from the 
discussion. 

TEST PROGRAM 

The objective of the test program was to determine the response of steel columns, base 
plates, and beam-to-column connections to blast loading and the effect of blast loading on 
the post-blast performance of connections. These tests consisted of four phases. In Phase 
I, the objective was to study the response of typical and blast-resistant base plate connec¬ 
tions to blast loading. Phase II was a study of the response of steel columns to blast load¬ 
ing. 


Protection Engineering Consultants 


1 



Report Documentation Page 


Form Approved 
OMB No. 0704-0188 


Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and 
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, 
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington 
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it 
does not display a currently valid OMB control number. 


1. REPORT DATE 

OCT 2010 


2. REPORT TYPE 

N/A 


3. DATES COVERED 


4. TITLE AND SUBTITLE 

Performance Of Blast-Damaged Steel Connections In Progressive 
Collapse 


6. AUTHOR(S) 


5a. CONTRACT NUMBER 


5b. GRANT NUMBER 


5c. PROGRAM ELEMENT NUMBER 


5d. PROJECT NUMBER 


5e. TASK NUMBER 
5f. WORK UNIT NUMBER 

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION 

Protection Engineering Consultants, Austin, TX report number 

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS (ES) 10. SPONSOR/MONITOR’S ACRONYM(S) 

11. SPONSOR/MONITOR’S REPORT 
NUMBER(S) 

12. DISTRIBUTION/AVAILABILITY STATEMENT 

Approved for public release, distribution unlimited 

13. SUPPLEMENTARY NOTES 

See also ADA550809. Military Aspects of Blast and Shock (MABS 21) Conference proceedings held on 
October 3-8, 2010. Approved for public release; U.S. Government or Federal Purpose Rights License., The 
original document contains color images. 

14. ABSTRACT 

In progressive collapse events initiated by explosive threats, damage to some connections from blast 
loading of the structure is likely. In this paper, the results from ten tests of blast-damaged and undamaged 
beam-to-column connections are examined. An evaluation procedure leading to comparisons of 
performance based on rotational ca-pacity of both blast damaged and undamaged capacity is presented. 


15. SUBJECT TERMS 


16. SECURITY CLASSIFICATION OF: 


a. REPORT 

unclassified 


b. ABSTRACT 

unclassified 


c. THIS PAGE 

unclassified 


17. LIMITATION OF 

18. NUMBER 

ABSTRACT 

OF PAGES 

SAR 

9 


19a. NAME OF 
RESPONSIBLE PERSON 


Standard Form 298 (Rev. 8-98) 

Prescribed by ANSI Std Z39-18 






In Phase III, six specimens representing four types of beam-to-column connections were 
subjected to blast loading. The connections joined spandrel beams to a single column, as 
shown in Figure 1, which is a plan view for the steel frames and reaction structure used 
for Phase III and IV. For all Phase III tests, concrete cladding was added to the frames to 
simulate in-service conditions for blast-pressure collection. 

In Phase IV, the blast-damaged connections from Phase III were loaded quasi-statically 
to failure. Measurements included the vertical displacement of the connection, down¬ 
ward load in the actuator, axial force in the spandrel beams, and strain in the spandrel 
beams. For comparison, four connections not subjected to blast loading were also tested. 
Therefore, Phase IV consisted of ten specimens, five types of connections. A summary 
of the connections tested in Phase III and IV is provided in Table 1. For reference, the 
table includes the ASCE 41-06 classifications for the connections (4). 


Table 1. Summary of Phase III and IV Tests 


Test 

No. 

Connection 

Type 

Beam 

Size 

ASCE 41-06 
Classification 

Test 

Phase 

1 

Traditional Seismic Moment 

W18x35 

fr 2 -wuf 3 

III, 

2 

Sideplate® Seismic Moment 

W18x35 


iii, 

3 

Coverplate Wind Moment 

W16x26 

FR-Welded Coverplated 

hi, 

4 

Traditional Seismic Moment 

W18x35 

FR-WUF 

hi, 

5 

Sideplate® Seismic Moment 

W18x35 


hi, 

6 

Traditional Seismic Moment 

W18x35 

FR-WUF 

IV 

7 

Sideplate® Seismic Moment 

W18x35 


IV 

8 

Coverplate Wind Moment 1 

W16x26 

FR-Welded Coverplated 

IV 

9 

Bolted Shear Tab 

W16x26 

PR 4 -Shear w/o Slab 

IV 

10 

Bolted Split Tee 

W18x35 

PR-Double Split Tee 

hi, 


Connection had extra transverse welds on both bottom coverplates joining the cover- 
plates to the bottom flanges of the spandrel beams. 

2 Fully restrained 

3 Welded Unreinforced Flange 

4 Partially restrained 


Protection Engineering Consultants 


2 















Figure 1. Plan View of Steel Frame for Phase III and Phase IV Tests (I) 


Protection Engineering Consultants 


3 




















































































EVALUATION PROCEDURE 


An evaluation procedure was applied to all ten Phase IV tests. The steps of this proce¬ 
dure are illustrated below for test no. 4, a traditional seismic moment (WUF) connection 
subjected to blast loading. 

The downward force applied to the column versus the Phase IV chord rotation is shown 
in Figure 2. The chord rotation was determined using the arctangent, as illustrated in 
Figure 3, because the beams reached relatively large rotations. For all connections, the 
initial Phase IV position of the connection was taken to be the post-blast final position 
from Phase III, i.e., a rotation of zero in Figure 2 corresponds with the chord rotation of 
the beam after blast loading. The rotations where connection articles (bolt or flange) 
failed are identified in Figure 2. The connection at ultimate failure is shown in Figure 4. 
The axial force in the spandrel beams was calculated using both load pins and strain 
gages, and the results of both calculations are shown in Figure 2. Because for most 
specimens, the load-pin results were more consistent than the strain-gage results, all con¬ 
clusions were based on the load-pin calculations. 



0.00 0.02 0.04 0.06 0.08 0.10 

Phase IV Chord Rotation [radians] 

Figure 2. Traditional Seismic Moment (WUF) Connection: 
Applied and Axial Force Versus Phase IV Chord Rotation 


Protection Engineering Consultants 


4 























L 



0: chord rotation of beam 


0 = arctan 


v-w 


Figure 3. Geometric Parameters of Frame Deformation 




Figure 4. Traditional Seismic Moment (WUF) Connection at Ultimate Failure (5) 


The moment at the connection was calculated by taking equilibrium of the spandrel beam 
about the connection. The moments calculated from both the load-pin and the strain-gage 
data are shown in Figure 5. Article failures are labeled, and the rotations where failures 
occurred are identified by vertical dashed lines. The initial failure (bottom flange and 
first bolt) takes place at 0.047 radians. Thereafter, the second and third bolts from the 
bottom fail at 0.060 and 0.078 radians, respectively. Finally, the top flange and fourth 
bolt from the bottom fail at 0.092 radians. The stiffness of the W18x35 beam is included 
for comparison. 


Protection Engineering Consultants 


5 













Phase IV Chord Rotation [radians] 

Figure 5. Traditional Seismic Moment (WUF) Connection: 
Moment Versus Phase IV Chord Rotation 


ROTATIONAL CAPACITY BY TYPE 

To characterize rotational capacity by connection type, 80%-failure rotations (80% of the 
rotation at first article failure) were calculated for each specimen and averaged for each 
connection type. For the traditional seismic moment (WUF) connection discussed above, 
the 80%-failure rotation was 0.80 x 0.047 = 0.038. This 80% value was taken to be a rea¬ 
sonable limiting rotation to develop the flexural capacity of the structure within a margin 
of safety. A larger rotation would increase the possibility of the first failure of a bolt or 
weld, which could introduce unpredictable loads into the structure. 

The results of these calculations are shown in Figure 6. The number of specimens for 
each connection type is shown on the average bar for that type; this convention is re¬ 
tained throughout this section, for all bar graphs. The minimum 80%-failure rotation of 
each type is also included, for comparison. Note that these specimens include both blast- 
damaged and pristine connections. 


Protection Engineering Consultants 


6 
















0.10 


^ 0.08 

c 

.2 

'~a 

E 

— 0.06 

Q) 

W> 

c 

< 

c 0.04 

o 

'+-* 

fU 

+-* 

o 

“ 0.02 


0.00 


Figure 6. Average and Minimum of 80%-Failure Rotation, 
by Connection Type 

As shown in Figure 6, the bolted shear tab and Sideplate® connections have the highest 
rotational capacities. The traditional seismic moment (WUF) and bolted split tee connec¬ 
tions have the lowest rotational capacities. The coverplate connection has an intermedi¬ 
ate capacity, with the minimum significantly less than the average. Many of the differ¬ 
ences between the coverplate specimens, noted here and below, were likely due to a fab¬ 
rication difference and blast damage and may not be inherent to the connection type. 
Specifically, the bottom flange of the beam joining the blast-loaded coverplate connec¬ 
tion was damaged in the Phase III test, weakening it for Phase IV; the coverplate connec¬ 
tion not subjected to blast-loading had an extra transverse weld (6). 

To assess the effect of blast loading on rotational capacity, 80%-failure rotations of blast- 
damaged and undamaged connections were compared. For two connection types—the 
traditional seismic moment (WUF) and Sideplate® connections—two specimens were 
subjected to blast loading. The rotations of these replicates were averaged and the results 
are shown in Figure 7. From the figure, blast loading decreased the rotational capacities 
of all connections but caused the largest decrease in the case of the coverplate connec¬ 
tion. 


■ Average 



tee tab 


Protection Engineering Consultants 


7 

























0.12 


to 

c 0.10 

ro 

‘-a 

E 

“ 0.08 

c 

o 

2 0.06 

2 

a; 

j 5 0.04 

LL. 

§ 0.02 

oo 

0.00 

B/asf Loaded 
Connection Type 



Figure 7. Effect of Blast Loading on 80%-Failure Rotation, 
by Connection Type 


CONCLUSIONS 

The test data and subsequent evaluations provide insight into the relative performance of 
the five connection types. The most important measure of performance is rotational ca¬ 
pacity because if a structure lacks ductile connections, it will likely lose structural conti¬ 
nuity, increasing the probability of progressive collapse. Additionally, ductile, fully re¬ 
strained moment connections can develop full member capacity over relatively large de¬ 
formations, can absorb significant amounts of energy associated with gravity induced ac¬ 
celerations and can reduce potential for collapse. The Sideplate® connection is a good 
example, as it exhibited the second largest rotational ductility of all the connections 
tested. Partially restrained connections exhibiting significant ductility (large rotations), 
while not contributing significant member capacity, can ensure continuity. The bolted 
shear tab had the largest rotational capacity, but the bolted split tee connection had rela¬ 
tively low rotational capacity. For the coverplate, Sideplate, and traditional seismic mo¬ 
ment (WUF) connections, blast loading was found to reduce rotational capacity. Finally, 
it bears noting that a limited number of tests were performed, some without replicates, 
and it is recommended that additional testing be performed on the connections discussed 
here as well as additional connections that are commonly employed. 


Protection Engineering Consultants 


8 




































REFERENCES 


1 Sheffield, Craig S. and Ford, Jeffrey S. Quick Look Data Report (Restricted Report, April 

2006). Defense Threat Reduction Agency, Kirtland AFB, NM. 

2 Sheffield, Craig S. and Ford, Jeffrey S. 1-6 Results Report (Restricted Report, July 2005). De¬ 

fense Threat Reduction Agency, Kirtland AFB, NM. 

3 Sheffield, Craig S. and Ford, Jeffrey S. 7-12 Results Report (Restricted Report, March 2006). 

Defense Threat Reduction Agency, Kirtland AFB, NM. 

4 Seismic Rehabilitation of Existing Buildings, ASCE Standard 41-06 (2006). American Society 

of Civil Engineers, Reston, VA 

5 Courtesy of Applied Research Associates, Albuquerque, NM 

6 Sheffield, Craig and Morrill, Ken. Phase III and IV Results and Calculation Summary 

(Restricted Report, February, 2008). Karagozian & Case, Albuquerque, NM 


Protection Engineering Consultants 


9