Title: Experimental Evaluation of the Seismic Performance of Piping Subassemblies
1Experimental Evaluation of the Seismic
Performance of Piping Subassemblies
- Presenter Dr. Emmanuel Manos Maragakis
- Dr. Ahmad Itani, Robert Corbin and Elliott Goodwin
2Personnel
- Authors
- Robert C. Corbin
- Elliot R. Goodwin
- Dr. Emmanuel Manos Maragakis (PI)
- Dr. Ahmad M. Itani (CoPI)
- Graduate Students
- Elliot R. Goodwin
- Robert C. Corbin
3Outline
- Motivation
- Objectives and Scope
- Experimental Program
- Input Motion Generation
- Experimental Protocol
- Experimental Results
- Experimental Comparisons
- Observations and Conclusions
4Motivation
- Why Minimize Nonstructural Damage?
- Life Safety
- Property Loss
- Direct Damage
- Consequential Damage
- Loss of Function
5Investment Costs
6Background/Objectives
- Background
- Experimental Evaluation
- Steel/Welded and Steel/Threaded Piping Systems
(Goodwin 2004) - Copper/Soldered Piping Systems (Corbin 2006)
- Objectives
- Understand the seismic behavior of braced and
unbraced piping systems - Identify the capacity characteristics
- Investigate and compare responses
7Scope
- Scope
- Experimental Specimen
- Generated Input Motions
- Experimental Protocol
- Displacement and Acceleration Responses
8Experimental Configuration
- Modeled after University of California, Davis
Medical Center piping system - Modified to accommodate dimensions and geometry
restrictions of testing facility - Consisted of 100 feet of 3 inch and 4 inch
diameter piping - Four typical valves including a y-strainer, a
check valve, and two gate valves - Two water heaters
- One simulated heat exchanger
- Heat exchanger and water heaters were anchored to
the shake table - System was braced and hung from a stationary frame
9Test Set-up and Specimen Configuration
10Plan View
11Elevation View
12Piping System Setup
13Bracing Details
Transverse Brace Longitudinal Brace
14Bracing Layout
15Instrumentation
29 Stringpot (SP) (20 stroke) Displacement
Transducers
16 Crossbow (xbow) Kinemetrics
(2g) Accelerometers
16Input Motion Generation
- ICBO AC156
- A synthetic motion must be generated that
- Meets a desired response spectra based on 1997
UBC - Build, hold and decay envelope of 5, 10 and 15
seconds, respectively - SIMQKE Program
- A maximum acceleration of 1 g
17Synthetic Input Motions SIMQKE
18Input Motion Generation
- Northridge Near-Fault Ground Motions were
utilized to perform dynamic analyses - Earthquakes were generated utilizing the Specific
Barrier Model (Wanitkorkul and Filiatrault) - A total of 4 cases, 2, 5, 10, 20, probability
of exceedance in 50 years - 2 components per case, Fault-normal (Large) and
Fault-parallel (Small) - 25 earthquakes simulated for each component
19Synthetic Input Motions OpenSees
20Experimental Protocol
- 4 steel systems tested (SIMQKE Motions)
- Welded braced
- Welded unbraced
- Threaded braced
- Threaded unbraced
- 2 copper systems tested (Northridge Motions)
- Braced
- Unbraced
- Subjected to increasing intensities of the
derived input motions in the principle axes, N-S
and E-W, as well as a biaxial excitation at 45
with respect to the principle axes
21Experimental Results
- Observations and Performance
- Steel/Welded Braced and Unbraced
- Steel/Threaded Braced and Unbraced
- Copper/Soldered Braced and Unbraced
- Instrument Comparisons
- Braced and Unbraced Subassemblies
- Copper/Soldered, Steel/Welded and Steel/Threaded
Braced Subassemblies - Copper/Soldered, Steel/Welded and Steel/Threaded
Unbraced Subassemblies
22Experimental Results Welded Braced System
- Two failed braces
- Cable portion only
- Occurred at highest input motion (100 SIMQKE E-W
) - Permanent displacement of braces
23Experimental Results Welded Unbraced System
- Two failed hanger rods
- Occurred at highest input motion (100 SIMQKE
Biaxial )
24Experimental Results Threaded Braced System
- Flanged connection to heat exchanger began to
leak at relatively low amplitudes - 50 of SIMQKE Generated Input Motion
- Leak formed at the vertical run above the heat
exchanger at - 70 of SIMQKE Generated Input Motion
- The testing protocol was stopped after the first
100 SIMQKE motion
25Flange Connection Failure
26Selected Instrument Response nv17 Steel Piping
Subassembly
nv17
27Experimental Results Copper
Braced Subassembly
- 2 Probability of Exceedance of the Northridge
Earthquake Ground Motion in the East/West
Direction at 100 Strength - Maximum displacement response 3.80 in
- Maximum acceleration response 2.64 g
- Experimental Observations
- Piping did not experience any damage
- Permanent displacement of several braces
- 3 inch piping risers permanently offset due to
damage experienced by water heaters
28Experimental Results Copper
Braced Subassembly
- Cable and Brace Damage (Brace B7)
Sag
Displacement
29Experimental Results Copper
Unbraced Subassembly
- 2 Probability of Exceedance of the Northridge
Earthquake Ground Motion in the North/South
Direction at 100 Strength - Maximum displacement response 11.14 in
- Maximum acceleration response 2.64 g
30Experimental Results Copper
Unbraced Subassembly
- Experimental Observations
- Displaced Hanger Rod
- 20 PE East/West Direction
- Drift of 2.3
- Hanger Rod Failure
- 10 PE North/South Direction
- Drift of 2.8
- Two Flange Failures
- 10 PE East/West Direction
- Drift of 2.6
31Experimental Results Copper
Unbraced Subassembly
32Experimental Results Copper
Unbraced Subassembly
- Failed Hanger Rod Connection
33Experimental Results Copper
Unbraced Subassembly
- Flange Connection Failure
34Experimental Comparisons Copper Braced vs.
Unbraced
- Displacement Responses
- Ranged between 1.12 to 16.8 times larger
- Acceleration Responses
- Ranged between 0.82 to 6.19 times larger
- Subjected to story drifts up to 3.13
- Braced subassembly suffered no damage
- Unbraced subassembly suffered minor damage
beginning at 2.3 and extensive damage at 2.6
35Response Summary
36Experimental Comparisons Copper Steel
Braced Subassemblies
- Experimental Observations
- Copper/Soldered
- Withstood a story drift of 3.13
- Steel/Welded
- Withstood a story drift of 5
- Steel/Threaded
- Leaks formed at a story drift of 2.17
- Failed in response to a story drift of 4.34
37Experimental Comparisons Copper Steel
Braced Subassemblies
- Displacement Responses
- Copper/Soldered produced the smallest
- Steel systems ranged between 1.42 and 13.1 times
larger - Acceleration Responses
- All three subassemblies were very similar
- Steel systems ranged between 0.57 and 4.00 times
larger
38Experimental Comparisons Copper
Steel Unbraced Subassemblies
- Experimental Observations
- Copper/Soldered
- Rod failure at 2.8 story drift
- Piping flanges failed at 2.6 story drift
- Steel/Welded
- Rod failure at 4.34 story drift
- Piping withstood a 4.34 story drift
- Steel/Threaded
- Minor joint leaks started at 1.08 story drift
- Joint failures at 2.17 story drift
39Experimental Comparisons Copper
Steel Unbraced Subassemblies
- Unbraced responses were inconsistent
- Displacement Responses
- Steel systems produced the largest longitudinal
responses - Ranged between 1.88 and 4.65 times larger
- Transverse and vertical responses were similar
- Acceleration Responses
- Longitudinal responses were similar
- Steel systems produced the largest transverse and
vertical responses - Ranged between 1.01 and 3.55 times larger
40Response Summary
- Longitudinal Responses of SP21 and xbow2
41Observations and Conclusions
- Piping systems were determined to be drift
sensitive - Braces significantly reduced system displacement
response - Braces did not reduce the system acceleration
response
42Observations and Conclusions, continued
- Copper system
- Braced
- Survived 3.13 drift
- Unbraced
- Flanges were the weak points
- Welded system
- Braced and Unbraced
- Survived 4.34 drift
- Threaded system
- Braced
- Complete failure at 4.34 drift
- Unbraced
- Leaks began at 1.08 drift
- Complete failure at 2.17 drift
- Fittings were weak point
43Observations and Conclusions, continued
- 1997 UBC code limit on story drift is 2.5
- Braced
- Copper/Soldered experienced no damage
- Steel/Welded experienced no damage
- Steel/Threaded would have leaks
- Unbraced
- Copper/Soldered experienced damage
- Steel/Welded experienced no damage
- Steel/Threaded would have leaks at half of UBC
requirement
44Thank YouQuestions?