SEISMIC PERFORMANCE OF REINFORCED CONCRETE SQUAT

WALLS WITH PERMANENT PVC FORMWORK SYSTEM

Devin Sauer 1 , Jose Centeno 2 , Carlos E. Ventura 3

ABSTRACT

The stay-in-place concrete forming technology made primarily from polyvinyl chloride (PVC)
has already shown improvements in durability, maintenance, construction time, and water-
tightness, when compared to traditional methods. In-plane dynamic testing was carried out on the
PVC formwork system with the objective to determine its performance under seismic loading.

Two reinforced concrete squat walls were built with the same amount of reinforcement, and only
one of the walls with the permanent PVC formwork system. The walls were modified by placing
openings in the middle to reduce their strength. A synthetic earthquake record, scaled to local
code, was used for both specimens. The amplitude of this record was increased incrementally
throughout the experiment and the results of the dynamic tests showed that, by comparison, the
structure with the permanent PVC formwork was protected against spalling and had a greater
lateral load capacity. In addition, there was a significant reduction of the amount of cracks on the
surface of the tested specimen.

1
Undergraduate Student, Univ. of British Columbia, Vancouver, BC, Canada,
email [email protected]

2
M.ASc. Student, Univ. of British Columbia, Vancouver, BC, Canada,
e-mail [email protected]

3
Prof. of Civil Engineering, Univ. of British Columbia, Vancouver, BC, Canada
email [email protected]

1
 INTRODUCTION

The stay-in-place polyvinyl chloride (PVC) formwork is designed to be easily constructed onsite
by connecting pre-manufactured PVC panels. When constructed, the formwork has finished
exterior sides and both horizontal and vertical divisions within. The formwork can be used in all
practical building designs of any size and is already available on the market.

A series of dynamic tests were conducted on two reinforced concrete squat walls with and
without the permanent PVC forming system, at the Earthquake Engineering Research Facility
(EERF) on April 10 and 16 of 2007. The prototype was subjected to prescribed simulated ground
motions at different intensities on a linear shake table in order to determine its performance under
severe dynamic loads in comparison to the bare specimen. The objective of these tests was to
determine whether the PVC forming technology protects a reinforced concrete structure against
cracking and spalling during a severe earthquake, and if the lateral load capacity of the prototype
exceeds that of the bare specimen.

BACKGROUND

Initially, the test specimens were designed as reinforced concrete squat walls with minimum
reinforcement [CSA, 2004] and a height to width ratio of 1.0. During testing, the maximum
flexural and shear cracking strength of the benchmark specimen surpassed the maximum dynamic
loads achievable by the newly installed EERF linear shake table. Consequently, structural
modifications on each wall were performed, by cutting an opening in the center, in order to
reduce the effective strength of the specimens. Because of this change, the specimens were not
detailed to have the reinforcement required at the corners of the opening.

TEST SPECIMEN DESCRIPTION AND SETUP

The wall dimensions were approximately 2.4m x 2.5m, with an opening in the center of 1.6m x
1.3m and a thickness of 0.10m. At the time of testing, the concrete strength was measured to be
32MPa and the yield strength of the reinforcement used was approximately 400MPa. The bare
minimum code amount (ρ=0.29) of horizontal and vertical reinforcement was provided [CSA,
2004]. Figure 1 shows the detailing of the reinforcement.

The test specimen foundation was bolted with high strength threaded steel rods to the EERF
shake table frame to achieve a rigid connection. Aluminum shims were used in areas in which
the base of the specimen foundation was uneven with respect to the table. To represent the
surcharge weight, ten steel plates, each having a weight 44.6kN were bolted to the slab of the
specimen.

Instrumentation

The instrumentation for this experiment consisted of piezo resistive accelerometers and position
transducers to measure the accelerations and the horizontal displacements of the specimen,

2
respectively. The accelerometers were fixed to the shake table, the specimen foundation, and the
top of the surcharge weight. Each accelerometer was hardware filtered to a high pass of 50Hz.
Two position transducers were placed at the top and at the base of the specimen to measure the
horizontal displacement at those locations. Table 1 provides a description and Figure 2 shows the
instrumentation used for these tests.

Table 1: Instrumentation Specifications

Channel
Instrument Type Range
ID

Piezo Resistive Accelerometer

Uniaxial Accelerometer A A1D-8 +/- 10 g
ICSensors Model 3110
B A3D-X +/- 5 g
Piezo Resistive Accelerometer
Triaxial Accelerometer C A3D-Y +/- 5 g
ICSensors Model 3026
D A3D-Z +/- 5 g
Celesco Cable-Extension
Position Transducer E SP13 Position Transducer +/-3000 mm
Model PT101-0150-111-1110

Celesco Cable-Extension
Position Transducer F SP19 Position Transducer +/-3000 mm
Model PT101-0150-111-1110

INTERFACE LoadCell
LoadCell G LoadCell +/- 50 kip
Model 1032-AFM
Piezo Resistive Accelerometer
Uniaxial Accelerometer H TAccel +/- 10 g
ICSensors Model 3110

PROCEDURE

The calculated natural frequency obtained through an analytical model using a finite element
method software, for specimen 1, with the surcharge weight of 44.6KN, resulted in Fn=18Hz. To
evaluate the performance of the test specimens, the selected record should generate similar
seismic demands in comparison to the design earthquake ground motion recommended by the
NBCC 2005 [NRCC, 2005].

Each test specimen was subjected to the VERTEQ-II synthetically generated earthquake
acceleration record [Telecordia Technologies, 1995], and the amplitude was consecutively
increased for each test done with the record. The VERTEQ-II record was normalized in order to
conform to the Uniform Hazard Spectrum for 2% probability of exceedance in 50 years for

3
Vancouver [Geological Survey of Canada, 2003], resulting in a peak acceleration of 0.50g. The
effective duration of the VERTEQ-II record was 16.0 sec and the sampling rate was 200 samples
per second (sps).

The tests were augmented by the addition of the records Llolleo (Valparaiso, 1985) and El Centro
(1940). The amplitude of the grounds motions were scaled to reach a PGA of 0.60g and 0.80g,
respectively. Furthermore, the duration of Llolleo (1985, Valparaiso) was 90 seconds and for the
El Centro (1940) record, the duration was 45 seconds. In addition, sequences of the VERTEQ-II
record were run with 30 seconds waiting time between records. Figure 3 shows the measured
ground motions at the base of the specimen for the different records and their corresponding
acceleration response spectrum.

RESULTS

Both specimens were loaded with the 44.6 kN surcharge weight 3 days prior to their testing. On
the testing day, an initial flexural crack was observed at the mid-span of the beam in the
benchmark specimen; whereas, the specimen with the PVC formwork remained uncracked.
Figure 4(b) shows the beam of the benchmark specimen on the day of testing, with the crack
traced in red.

Specimen #1 (Benchmark Test Structure)

Table 2 shows the loading sequence that specimen one underwent.

Table 2: Loading Sequence for Specimen 1

Test # Description
1 VERTEQ-II 25%
2 VERTEQ-II 50%
3 VERTEQ-II 75%
4 VERTEQ-II 100%
5 VERTEQ-II 200%
6 Llolleo
7 El Centro
8 A sequence of two VERTEQ-II 100% records
9 A sequence of two VERTEQ-II 200% records

At amplitudes of 50% and 75% of the VERTEQ-II record, diagonal hairline cracks developed at
the corner of the openings. At 200%, the test specimen sustained dynamic loads above its
maximum strength and reached severe cracking. Flexural cracks developed along the length of
the columns, with the largest flexural cracks occurring at the ends of the clear span. Additional
flexural cracking was observed in the beam along mid-span and at the ends of the clear span. The
width and length of existing diagonal cracks from previous dynamic loading were significantly
increased, and were observed at all four corners of the opening. All flexural cracks along the
span of the beam and columns coincided with the location of the vertical and horizontal
reinforcement, respectively. Figure 4(c) and 4(d) show the crack distribution in the column and
the beam respectively.

4
The completion of tests 6, 7, and 8 showed existing cracks opening further, but no new cracks
appearing. The collapse of the benchmark specimen occurred during the second record in the
200% sequence (test 9). A crack developed due to a combination of flexure and shear at the
bottom of the clear span of the left column; consequently, allowing significant rotation. When
the subsequent spalling of the concrete cover occurred, this column became unstable and the
specimen collapsed out of plane.

Specimen #2 (Structure with permanent PVC formwork)

Table 3 shows the dynamic loading progression for the structure with the PVC formwork.

Table 3: Loading Sequence for Specimen 2

Test # Description
1 A sequence of four VERTEQ-II records 25%, 50%, 75%, 100%
2 VERTEQ-II 200%
3 VERTEQ-II 225%
4 VERTEQ-II 250%
5 Llolleo
6 El Centro
7 A sequence of two VERTEQ-II 100% records
8 A sequence of three VERTEQ-II 200% records
9 A sequence of three VERTEQ-II 225% records
10 A sequence of three VERTEQ-II 250% records

When the VERTEQ-II record was run at the maximum level achieved during the benchmark test
(200% amplitude) and no discernible damage was observed, it was decided that the amplitude of
the record should continue to be incrementally increased until cracking of specimen 2 occurred.

At an amplitude of 225%, a crack was observed in the upper right corner of the opening, which is
shown in Figure 5(c). To observe the state of this concrete segment, a small portion the PVC
form was removed, which exposed a diagonal crack starting at the corner of the opening and
stopping at the first vertical division of the panel.

When the consecutive test was done (VERTEQ-II 250%), the diagonal crack previously observed
opened up and increased in length. Also, Figure 5(d) shows a hairline flexural crack which
formed on the right corner of the beam.

To better observe the crack width under the PVC cover, some of the material was partially
removed from all four corners of the opening after tests 5 and 6. The results showed that at the
corners of the opening and at the ends of the beam, existing cracks opened further and additional
hairline flexural cracks formed along the span of the beam.

The final tests done on the specimen produced no additional cracks; however, existing cracks
became wider and longer. The state of specimen 2 at the end of testing showed a minimal
amount of diagonal cracks at the corners of the opening and considerable flexural cracks at the
ends of the beam. All flexural cracks observed along the span of the beam coincided with the
location of the center line of each PVC panel.

5
Comparison of test results for both specimens

Prior to testing the benchmark specimen showed hairline flexural cracks at the mid-span of the
beam due to the applied surcharge weight; whereas, no flexural cracks were observed for the test
specimen with the PVC formwork. The incremental dynamic loading caused the benchmark test
specimen to undergo severe flexural cracking and eventually spalling at the lower end of the left
column. The test specimen with the PVC formwork showed no flexural cracks on the columns
after partially removing the panels from the surface at the same location.

Flexural cracks formed along the span of the beam for both specimens although, the location of
cracks significantly differed. For specimen 1, flexural cracks coincided with the location of the
vertical reinforcement. For specimen 2, flexural cracks mostly coincided with the location of the
center line of each PVC panel. The widths of the cracks in the beam of specimen 2 were smaller
than the corresponding cracks observed in specimen 1.

Figure 6 shows the variation of the inertial force versus the displacement for both specimens
during the dynamic loading at VERTEQII 200%. The load deformation plot for specimen 2
shows a higher stiffness than specimen 1, which relates to difference in cracking observed for
both specimens at this stage of testing. After increasing the dynamic load to VERTEQII 250%,
the load deformation plot for specimen 2 showed a similar stiffness to specimen 1 at VERTEQII
200% (see Figure 6b). This similar decrease in stiffness for both specimens indicates cracks were
also formed within the cross section of the specimen with the PVC formwork.

CONCLUSIONS

1. For the test specimen with the PVC formwork, no flexural cracks were found in the columns
after partially removing the panels.
2. Flexural cracks formed along the span of beam for both specimens; however, the location of
cracks significantly differed.
3. The width of the cracks on the beam of specimen 2 were smaller than the corresponding
cracks on the benchmark specimen.
4. Instrumentation measurements indicated that during the increase of dynamic loading, the loss
in stiffness for specimen 2 was similar to that of the benchmark specimen, but for a higher
level of dynamic loading. This suggests that cracking may have occurred within the cross
section of specimen 2.
5. Specimen 2 in comparison to specimen 1 had a greater lateral load capacity and could resist
more cycles of shaking.
6. The PVC formwork protected a reinforced concrete structure against spalling during a severe
earthquake.

RECOMMENDATIONS

Cyclic testing should be performed on different reinforced concrete structures with the PVC
formwork, measuring the strains in the different materials to establish the formwork’s influence
on the displacement and strength capacity of a structure.

6
 REFERENCES

CSA, 2004. Design of concrete structures. Committee A23.3, Canadian Standards Association,
Toronto, Ont.

NRCC, 2005 “National Building Code of Canada, 2005 Part 4 Structural Design”

Geological Survey of Canada, 2003 “Fourth generation seismic hazard maps of Canada: Values
for over 650 Canadian localities intended for the 2005 National Building Code of Canada”

Telcordia Technologies,“Network Equipment-Building Systems (NEBS) Requirements: Physical

Protection”. Telcordia Technologies Generic Requirements, GR-63-CORE, Issue 1, October
1995.

7
 φ 10mm φ 10mm @125mm

Vertical reinforcem ent

φ 10mm @300mm

Horizontal reinforcement
φ 10mm @300mm

A A
Lap Splice
400mm
Lap Splice
450 mm

Front View

Section A-A
Hairpin φ 10mm @ 300mm

Figure 1: Test Specimen Dimensions

B, C, D

G, H

Figure 2: Instrumentation

8
VERTEQ-II (100%)
Acceleration (g)
Acceleration Response Spectrum ξ =5%
1 3.0
0.5
0 2.5 Measured Fbk
0.5 UHS 2% in 50 YRS
1 2.0
0 5 10 15 20 25 30 35 40

PSA (g)
Time (s) 1.5

Displacement (mm) 1.0

200
100 0.5
0
100 0.0
200
0 5 10 15 20 25 30 35 40 0.1 0.5 1 5 10 50 100

Time (s) Frequency (Hz)

LLOLLEO
Acceleration (g)
Acceleration Response Spectrum ξ =5%
2 4.0
1
3.5 Measured Fbk
0
1 3.0
2
0 10 20 30 40 50 60 70 80 90 2.5
PSA (g)

Time (s) 2.0

Displacement (mm) 1.5

500 1.0
250
0 0.5
250 0.0
500
0 10 20 30 40 50 60 70 80 90 0.1 0.5 1 5 10 50 100

Time (s) Frequency (Hz)

EL CENTRO
Acceleration (g)
Acceleration Response Spectrum ξ =5%
2 9.0
1 8.0
0 Measured Fbk
1 7.0
2 6.0
0 5 10 15 20 25 30 35 40 45
PSA (g)

5.0
Time (s)
4.0
Displacement (mm) 3.0
500 2.0
250
0 1.0
250 0.0
500
0 5 10 15 20 25 30 35 40 45 50 0.1 0.5 1 5 10 50 100

Time (s) Frequency (Hz)

Figure 3: Ground Motions and Response Spectra

9
 Prior to Testing
(a) (b)

After VERTEQ-II at 200%

(c) (d)

Collapse
(e) (f)

Figure 4: Test Photos for Specimen 1

10
 Prior to Testing
(a) (b)

After VERTEQ-II at 225% After VERTEQ-II at 250%

(c) (d)

After Final Test

(e) (f)

Figure 5: Test Photos for Specimen 2

11
 Dynamic Loading #5
VERTEQ-II (200%)
160
Specimen 1
120
Specimen 2
Inertial Force [kN]

80
40
0
-40
-80
-120
-160
-20 -16 -12 -8 -4 0 4 8 12 16 20
Relative Displacement [mm]
(a)
Dynamic Loading #7
VERTEQ-II (250%)
160
120 Specimen 2

80
Inertial Force [kN]

40
0
-40
-80
-120
-160
-20 -16 -12 -8 -4 0 4 8 12 16 20
Relative Displacement [mm]
(b)
Figure 6: Inertial Force vs. Displacement for
a) VERTEQII 200%
b) VERTEQII 250%

12