Assessment of

Existing Structures
Using Cyclic Load
Testing
Case studies illustrate procedures recommended in ACI Committee 437 report

By Tarek Alkhrdaji, Nestore Galati, and antonio nanni

W hen a building is renovated for a change of use, the

load-carrying capacity of the structural system must
be established. Load testing can be used to provide
reducing the risk of overloading or damaging the
structure. Using hydraulic rams may also be more
economical than using weights. Although hydraulic
reliable verification that a given structure can safely rams require reaction systems that can be expensive
support the calculated design loads. and time consuming to implement, the labor required to
Per ACI 318,1,2 the test load magnitude (TLM) is required apply gravity loads using dead weights can also be
to be reached in at least four load increments. A set of expensive. The CLT method is discussed in greater
response measurements (mainly deflection) is taken after detail in ACI 437R4 and ACI 437.1R.5
the total test load has been applied and after at least In the following case studies, the CLT method was
24 hours of sustained loading. A final set of response used to verify analyses and capacities of existing structural
measurements is also required 24 hours after the test load components. The method was also used to evaluate
has been removed, so the total duration of the load test structural behavior after members were strengthened
can exceed 72 hours. When multiple load tests are required using externally bonded fiber-reinforced polymer (FRP)
to verify the capacities of multiple elements or configurations, reinforcement alone or in conjunction with a reinforced
significant delays and expenses can be incurred. concrete topping slab.
In the past 10 years, researchers and practitioners in
the U.S. have been evaluating an alternative load test Case studies
method.3 Known as the cyclic load test (CLT) method, For both of the structural investigations described
this procedure requires the application of multiple herein, a building floor was evaluated for a change in use.
cycles of loading and unloading (typically six). Structural Following industry recommendations, existing conditions
adequacy is then verified by examining the linearity were assessed by studying existing drawings, reports,
of the measured deflection response and magnitude and calculations; and the information was verified using
of the permanent deformation after the load has been on-site inspections.4,6,7 In each case, the assessments
removed. showed that strengthening would be required. Also, in
CLT investigations are typically conducted using each case, the CLT method was selected to minimize the
hydraulic rams that allow the test member to be quickly duration of the testing program, as testing was required
unloaded at any sign of distress, improving safety and both before and after strengthening.

Concrete international / april 2010 39

 Per ACI 437.1R-075 recommendations, loading and shelving system to store microforms collections. The
shoring systems were designed to ensure safety, prevent elevated slab was originally designed for 100 psf (4.7 kN/m2)
collapse of the test member, and avoid damaging adjacent live load. The filing system required that the slab be
structural elements. Also per ACI 437.1R-07,5 acceptance upgraded to a live load demand of 200 psf (9.6 kN/m2).
criteria—including deflection repeatability, permanency, During the site investigation phase, flexural cracks
and deviation from linearity—were used to examine the were observed on the top of the slab. The discovery
performance during and after the load test (Fig. 1). prompted NIH officials to request load tests to verify the
capacity of the existing floor slab prior to strengthening.
National Institutes of Health Library The load test was designed to induce loads mimicking the
Level B of Building 38 at the National Institutes of Health original design loads on certain areas of the slab.
(NIH) in Bethesda, MD, houses the National Library of Level B comprises a 10.5 in. (265 mm) thick concrete
Medicine. This level was being renovated to accommodate flat plate slab reinforced with 40 ksi (275 MPa) deformed
a new high-density filing system and a new carousel steel bars. A typical bay is supported by 24 x 34 in.
(610 x 860 mm) reinforced concrete columns on a 21 x 21 ft
(6.4 x 6.4 m) grid. The CLT was conducted on a 10.5 ft
Index Calculation* Limit (3.2 m) wide column strip located along Grid Line 12 to
2 2 evaluate the current bending capacity at midspan and at
Repeatability, ∆ max −∆r
IR I R= 1 1 0.95 ≤ IR≤ 1.05 the support (Fig. 2).
∆ max −∆r
The TLM was determined using ACI 437.1R-075
Permanency, ∆ r2
IP= IP ≤ 0.10
IP 2
∆ max
TLM = Dw + 1.1 Ds + 1.6L Eq. (1)
Deviation from Linearity,
IDL IDL = 1− [
tan α i
tan α ref [ IDL< 0.25
where Dw is the dead load due to slab self-weight (130 psf
* Measured values used in calculations: [6.2 kPa]), Ds is the sum of the superimposed dead loads
∆2max = maximum deflection in Cycle 2 under a load of Pmax (25 psf [1.2 kPa]), and L is the specified live load (125 psf
∆2r = residual deflection after Cycle 2 under a load of Pmin [6.0 kPa]) per the original design.
∆1max = maximum deflection in Cycle 1 under a load of Pmax The test team also decided to perform two additional
∆1r = residual deflection after Cycle 1 under a load of Pmin loading cycles using the load magnitude per ACI 318-05,
= maximum load level achieved by Cycles 1 and 2 Chapter 20.1 For this case, the TLM was
Pmax
P = minimum load level achieved at the end of Cycles 1 and 2
min
TLM = 0.85 [1.4(Dw + Ds ) + 1.7L] Eq. (2)
tan α i = slope of secant line on load deflection envelope
tan α ref = slope of secant line for peak on first loading cycle
Using Eq. (1) and (2), the TLM values were quite
Fig. 1: Acceptance criteria per ACI 437.1R-075 similar—358 psf (17.1 kN/m2) and 365 psf (17.5 kN/m2),
respectively. Table 1 summarizes the moment and punching
shear capacities, fMn and fVn, and
the factored bending and punching
11 12 11 12
13 13 shear demands, Mu and Vu, for the
column strip under investigation
based on the as-designed (original)
J J
conditions. To perform the load test,
100 in
in. Load points concentrated loads P were applied
using hydraulic rams to mimic the
Load line 1
90 in. effect of the uniformly distributed
H H design loads on the test slab strip.
100 in. Load points
108 in. Critical cross section
Two loading configurations were
Load line 1
Load line 2 used to test the slab: Scheme 1 was
72
72 in.
in Critical cross section
Load line 2
used to reproduce the negative
bending at Column H12; and Scheme 2
G G
was used to reproduce the positive
(a) (b) bending at midspan between
Column G12 and H12, as shown in
Fig. 2: Load point locations used for evaluation of Level B for the National Institutes of
Health project: (a) Scheme 1, positive moment test; (b) Scheme 2, negative moment Fig. 2. Table 2 gives the magnitude
test (1 in. = 25.4 mm) of P for these load tests.

40 april 2010 / Concrete international

 As shown in Tables 1 and 2, the down load test was selected (Fig. 3), cycle consisted of loading the slab in
existing slab has higher shear using the deadweight of the floors a minimum of four approximately
capacity than that corresponding to above to resist reactions from the equal loading steps, followed by at
the new demand. Accordingly, the hydraulic rams. least two unloading steps. The
test load layout was intended to Once the load test components maximum load reached in Cycles 5
produce the target bending moments and instruments were installed, a and 6 corresponded to the load
without necessarily achieving the preliminary load of 3000 lb (13.3 kN) combination determined per Eq. (1),
shear force demand simultaneously. was applied to eliminate slack in the whereas the maximum load in Cycles 7
Additionally, for the Scheme 2 load load system. The slab was then and 8 was per Eq. (2). Table 3 gives
test, it was not possible to apply the tested using eight loading-unloading the load levels used in each cycle.
load symmetrically with respect to cycles for each test configuration, Results of the Scheme 1 test
Column H12 due to the presence of including four loading levels with two indicate a fairly linear behavior for
piping at those locations. A push- cycles for each load level. Each load positive moments. Repeatability,
permanency, and deviation from
linearity were within the limits
Table 1: prescribed by ACI 437.1R-07.5
Capacities and demands for existing structure Additionally, no new cracks were
fMn, Mu, fVn, Vu, observed while performing the cyclic
Test kip-ft kip-ft kip kip load test. Although existing cracks
scheme (kN-m) (kN-m) (kN) (kN) Objectives did widen during loading, they
returned to their original widths at
Evaluate performance of the end of the tests. Figure 4 shows
120.7 98.8 105.6 67.7
1 column strip at positive the applied load cycles for the
(163.7) (134.0) (469.7) (301.1)
moment region Scheme 2 test. As indicated in Table 3,
the limits on deviation from linearity
Evaluate performance of were not met in the last two cycles.
212.5 211.6 105.6 67.7
2 column strip at negative However, because no sign of failure,
(288.2) (286.9) (469.7) (301.1)
moment region such as excessive deflection or cracking,
was observed, the performance of
the structure was deemed acceptable.
Table 2: The load test results and the
Concentrated load values P used to mimic moment and shear effects of preexisting top-side cracks were
uniform TLM indications that the structure could
P, Mu(TLM), Mu(P), Vu(TLM), Vu(P), have been subjected to loads that
Test kip kip-ft kip-ft kip kip exceeded its original design live load
scheme (kN) (kN-m) (kN-m) (kN) (kN) of 100 psf (4.7 kN/m2). During further
14.1 93.3 93.5 64.2 46.2 investigation of the building’s loading
1 history, it was revealed that the floor
(62.7) (126.5) (126.8) (285.6) (205.5)
was used to shore the floor above
28.4 190.6 191.4 64.2 46.2 during a previous renovation. This
2
(126.3) (258.5) (259.5) (285.6) (205.5) might have overloaded the slab.

Table 3:
Acceptance evaluation for Scheme 2 results
Deviation from
Load Repeatability Permanency linearity
cycles Load level (95 to 105%), % (≤ 10%), % (≤ 25%), % Performance
1 and 2 D + Ds + L 104.3 3.9 2.8 Satisfactory
3 and 4 0.75 (1.0Dw + 1.1Ds + 1.6L) 103.9 3.3 12.3 Satisfactory
4 and 6 (1.0Dw + 1.1Ds + 1.6L) 101.7 9.2 24.6 Satisfactory
7 and 8 0.85 (1.4[Dw + Ds] + 1.7L) 104.1 9.9 26.5 Acceptable

Concrete international / april 2010 41

 12 11
Analytical predictions were based on a two-dimensional
14 13
finite element model using commercial software (SAP
2000). The model consisted of one-dimensional “beam
Shoring elements” representing existing columns and a fine mesh
of “plate elements” to represent the floor slab. The
concrete was assumed to be isotropic and linear elastic,
Steel plate Load and the modulus of elasticity was determined per ACI 318-08.2
cell Timber
Hydraulic block Slab cracking during the test was introduced into the
ram model by reducing the stiffness of the slab to the
Steel
Level B post effective stiffness as defined in Section 9.5 of ACI 318-05.1
Figure 5 compares the analytical predictions with the
Shoring experimental results for the Scheme 2 test. This figure
tower
below shows that deflections measured in the first two cycles
each
load matched those predicted for an uncracked slab, whereas
Level A point
the measured deflections in the last two cycles are closer
to deflections predicted based on a cracked slab condition.
Fig. 3: CLT setup for the National Institutes of Health project. A transitional behavior can be observed on the third to
Shoring was used to distribute the reaction forces to floors
above the test floor sixth cycles, indicating that as the test load increased,
cracks developed in the slab bringing the behavior close
to that of a cracked slab at the higher load levels.
30 140
0.85[1.4(D + Ds) + 1.7L To accommodate the new design load for the Level B
1.0D + 1.1Ds + 1.6L 120
floor, externally bonded carbon FRP was used to increase
25
the bending capacity of the slab. FRP strips were installed
100 in two directions on the top and bottom sides of the slab.
20
Design and detailing of the FRP were performed according
L o a d , kN

D + Ds + L
L o a d , kip

80 to ACI 440.2R8 guidelines.

15
60
Commercial retail building
10
40 To address the needs of a potential tenant, the owner
of a commercial building in Cleveland, OH, evaluated
5
20 options for upgrading the second level floor to house
telecommunications equipment. The live load required
0 0
0 5000 10000 15000 20000 25000 30000 35000
for this equipment ranged from 125 to 150 psf (6.0 to
T im e, seconds 7.2 kN/m2).
Fig. 4: Load cycles for Scheme 2 test on the National Institutes of The nine-story building was constructed in 1917 with
Health project. Load includes weight of loading apparatus a masonry skin on a concrete-encased steel frame and
reinforced concrete (RC) floor system. The existing
drawings provided only floor plans and geometry of the
0.05 1.4
members, but no details were available for the structural
E xperim ental R e su lts
1.2
steel members or steel reinforcement. The typical floor
F E M U ncra cked
0.04 system consists of 6 in. (150 mm) wide reinforced
F E M C racked
concrete joists supporting a 3.5 in. (90 mm) concrete
D e fle ctio n s, m m

1
D eflections, in.

slab reinforced with No. 3 (No. 10) bars spaced at 18 in.

0.03
0.8 (460 mm) on center. A typical joist has a total depth of
15.5 in. (40 mm) and a span of 27.6 ft (8.4 m), and the
0.6
0.02 joists are 26 in. (660 mm) on center.
0.4 Dimensions of the existing joists were field verified.
0.01 Condition assessment and site investigation revealed that
0.2 the joists were typically reinforced with two 1 in. (25 mm)
square bottom bars at midspan. About 5.5 ft (1.7 m) from
0 0
1 2 3 4 5 6 7 8 each support, one bar is bent up and extends as a top bar
Load C ycles over each support and into the adjacent span. An additional
Fig. 5: Comparison of results for Scheme 2 test on the National 1 in. (25 mm) straight top bar was located over the support
Institutes of Health project at each end of a joist. No transverse reinforcement was

42 april 2010 / Concrete international

 Full span saw cuts to
isolate test joists Load Beam

Test Joist 1 Load app lication Test 1

points and
Test 2 Timber
Pads Spreader
Test Joist 2 Beam

Test 3
Test Joist 3

H ig h
Fig. 6: Plan of the load test area for commercial building retrofit S tre n g th
project S te e l B a r
Load
C e ll
Reaction
H yd ra u lic
Beam
Ram

R C M icro
Pile

Fig. 7: Loading points for two joists test for commercial building Fig. 8: CLT setup for Tests 1 and 2 of the commercial building
retrofit project retrofit project

located in the joists. Based on available historical data and before and once after they were strengthened (Tests 1
observed conditions, a nominal concrete strength of 4000 and 2, respectively). A third joist was tested after it was
psi (27.6 MPa) and steel yield strength of 33 ksi (240 strengthened to examine the performance with no
MPa) were used for preliminary analysis of the joists. pre-induced damage (Test 3). The test joists were
The proposed new loads included a superimposed isolated by saw cutting the concrete slab to eliminate
dead load of 25 psf (1.2 kN/m2) for a new concrete overlay load sharing with adjacent members (Fig. 6).
to level the slab surface and a service live load of 150 psf Analytical modeling indicated that the maximum
(7.2 kN/m2). Analyses indicated that the joists were moment and shear forces due to the design uniform loads
deficient in both flexure and shear for the proposed could be replicated using two point loads, each located
loads, with an existing live load capacity of approximately 3 ft (0.9 m) from the joist midspan (Fig. 6 and 7). A
96 psf (4.6 kN/m2) governed by the shear strength of the pull-down-type load was used in these load tests. The load
existing joists. The existing shear capacity of a typical was applied using hydraulic rams that were connected to
joist was estimated at 9 kips (40 kN), whereas the shear (and pulled against) a reinforced concrete micropile
demand for the new load was approximately 11.4 kips installed at the ground floor (one level below) to provide
(50.7 kN). necessary reactions (Fig. 8). In each test, the load was
To eliminate the possibility of brittle shear failure, all applied in six cycles comprising two cycles at each of
test joists were strengthened for shear using externally three loading levels.
bonded CFRP prior to testing. Three load tests were Test 1 was performed on two joists isolated by saw
performed to verify the existing load-carrying capacity, cutting the concrete slab at mid-distance to the first
controlling failure mode, and strength improvement after adjacent joist on each side of the test joists. The two
the strengthening systems were installed. In Tests 1 and joists were then strengthened for shear with an externally
2, two joists were simultaneously load tested—once bonded CFRP system. Corners on the joist stems were

Concrete international / april 2010 43

 0 0

-0.2 -5

Deflection, mm
Deflection, in.
-0.4 -10

-0.6 Test 1 -15

Test 2
Test 3
-0.8 -20
0 0.125L 0.25L 0.375L 0.5L 0.625L 0.75L 0.875L 1.0L
Distance From Support

Fig. 9: CFRP shear strengthening of test joists of the commercial Fig. 10: Comparison of test results commercial building retrofit
building retrofit project project. Values were measured at P = 14,500 lb (64.5 kN)

rounded to a 0.5 in. (13 mm) radius to prevent stress regions). The number and distribution of the cracks
concentrations, and 12 in. (305 mm) wide strips of indicated that sufficient bond existed between the
U-wrap CFRP were installed at 16 in. (406 mm) spacing existing slab and new overlay to transfer horizontal shear
along the full span of each joist (Fig. 9). The calculated forces and produce monolithic behavior. The reinforced
shear strength of the joists with this CFRP configuration concrete overlay enhanced the strength and stiffness of
was 13 kips (58 kN). the test joists and reduced deflection (Fig. 10). Based on
Deflections and crack widths were monitored in the test results, the strengthened joists were rated as
real-time during the load test. Test 1 was terminated adequate to support their self-weight, a 36 psf (1.7 kN/m2)
when the midspan deflection indicated inelastic behavior. superimposed dead load (reinforced concrete overlay),
Failure of the joists was governed by yielding of reinforce- and 150 psf (7.2 kN/m2) live load.
ment at the support, as evidenced by a large crack that Test 3 was performed on a single joist that was
developed on the top side of the slab. The width of this isolated by saw cutting the slab on each side (Fig. 6).
crack increased until the load test was terminated. There Prior to testing, the joist was strengthened for flexure
was no indication of failure at midspan, as midspan crack using a bonded RC overlay and for shear using CFRP
widths were stable at maximum load. strips. To expedite the construction schedule and
Based on the results of Test 1, it was concluded that minimize construction cost, the CFRP layout for Test 3
the CFRP-strengthened joists were able to support a comprised vertical strips applied only to the sides of the
superimposed dead load of 25 psf (1.2 kN/m2) plus a live joist stem, thus avoiding the need to round the corners
load of 135 psf (6.5 kN/m2). The shear performance was of the joists. In addition, the system provided full
adequate with no shear cracks observed on the joist after coverage of the side faces of the joists, as the fibers in
the test was completed. one ply of the CFRP strip had fibers oriented in the
To resolve the observed negative bending deficiency, a vertical direction. The calculated shear capacity of the
bonded concrete overlay, approximately 3 in. (76.2 mm) strengthened joist was 14 kips (62.3 kN). The purpose of
thick and reinforced with a steel wire mesh, was installed Test 3 was to verify that this optimal CFRP layout would
on the same two joists after roughening the slab surface provide adequate shear performance and to examine
to approximately 0.25 in. (6 mm) amplitude. Test 2 was the performance of a strengthened joist that was not
performed after the concrete overlay cured. The joists previously damaged by load testing (as was the case for
were loaded cyclically, following the same protocol as Test 2 joists).
Test 1 but using a maximum test load of 85% of the In Test 3, the joist was loaded cyclically to 85% of the
factored design loads, as specified by Chapter 20 of ACI design factored loads. The strengthened joist had improved
318-05.1 This load level would not cause excessive stiffness relative to the damaged and strengthened joists
damage to the upgraded joists, thus eliminating the need evaluated in Test 2. The strengthened joist also had
for additional repairs after the test. almost twice the stiffness of the unstrengthened joist
As the load approached the maximum test load, a evaluated in Test 1. Based on the acceptance criteria
number of flexural cracks developed on the top side of parameters, the performance of the joist was considered
the overlay at both ends of the joists (negative moment satisfactory. As with the previous tests, no shear cracks

44 april 2010 / Concrete international

were observed in Test 3, confirming the adequacy of the 4. ACI Committee 437, “Strength Evaluation of Existing Concrete
alternate CFRP layout. Buildings (ACI 437R-03),” American Concrete Institute, Farmington
Hills, MI, 2003, 28 pp.
Change in use 5. ACI Committee 437, “Test Load Magnitude, Protocol and
In both of the described cases, the CLT method Acceptance Criteria (ACI 437.1R-07),” American Concrete Institute,
efficiently verified the capacities of the existing structures. Farmington Hills, MI, 2007, 38 pp.
For the first case, load testing was used to determine the 6. ACI Committee 364, “Guide for Evaluation of Concrete
load-carrying capacity of the existing slab, verify the Structures before Rehabilitation (ACI 364.1R-07),” American
cause of existing cracks, and confirm the reliability of the Concrete Institute, Farmington Hills, MI, 2007, 22 pp.
analytical models that were later used to determine the 7. SEI-ASCE Committee 11, “Guideline for Structural Condition
required level of strengthening at various locations. For Assessment of Existing Buildings,” (SEI-ASCE 11-99), American
the second case, the load tests provided information on Society of Civil Engineers, Reston, VA, 2000, 147 pp.
the load-carrying capacity of the existing joists and their 8. ACI Committee 440, “Guide for the Design and Construction of
governing failure mode, confirmed the performance and Externally Bonded FRP Systems for Strengthening Concrete
composite behavior of the bonded reinforced concrete Structures (ACI 440.2R-02),” American Concrete Institute, Farmington
overlay upgrade solution, and allowed for optimizing the Hills, MI, 2002, 45 pp.
shear strengthening solution using an externally bonded
CFRP system. Received and reviewed under Institute publication policies.
It should be emphasized that externally bonded CFRP
reinforcement provided a cost-effective strengthening
solution in both cases. Also, because load testing verified
that the existing structural components had adequate
capacity to carry the design service loads without the
contribution of the FRP, no additional fire protection was
needed for the CFRP. Only an intumescent top coat was
used to provide the smoke-density and flame-spread
ACI member Tarek Alkhrdaji is an
ratings required per the governing building codes.
Engineering Manager with the Strengthening
Division of Structural Group, Inc., Hanover,
Future considerations
MD. He is a member of ACI Committees 437,
Increased use of load testing can be anticipated as
Strength Evaluation of Existing Concrete
more owners opt to update rather than replace existing
Structures; 440, Fiber Reinforced Polymer
buildings. The CLT method offers significant time and
Reinforcement; and 562, Evaluation,
costs savings relative to the current ACI 3182 procedure,
Repair, and Rehabilitation of Concrete
but there is a need to understand the effects of short-term
Buildings. He is also a member of the
creep on both old and new structures when loaded to
International Concrete Repair Institute.
near their capacities. It would appear that this can be
achieved by using the two test methods to evaluate
Nestore Galati is a Design Engineer with
similar structural members and comparing the resulting
the Strengthening Division of Structural
deflection behaviors and residual deformations. Buildings
Group, Inc., Hanover, MD. He is a member
that are to be replaced can also be used for comparative
of ACI Committees 437, Strength Evaluation
tests to failure.
of Existing Concrete Structures, and 440,
Fiber Reinforced Polymer Reinforcement.
References
1. ACI Committee 318, “Building Code Requirements for Structural
Concrete (ACI 318-05) and Commentary (318R-05),” American
Concrete Institute, Farmington Hills, MI, 2005, 430 pp.
Antonio Nanni, FACI, holds professorships
2. ACI Committee 318, “Building Code Requirements for Structural
at the University of Miami and University
Concrete (ACI 318-08) and Commentary,” American Concrete
of Naples—Federico II. He is an active
Institute, Farmington Hills, MI, 2008, 473 pp.
member of several ACI technical committees.
3. RILEM Technical Committee 20-TBS, “General Recommendation
for Statistical Loading Test of Load-Bearing Concrete Structures In
Situ,” RILEM Technical Recommendations for the Testing and Use of
Construction Materials, E&FN Spon, London, England, 1994,
pp. 379-385.

Concrete international / april 2010 45