Repair of columns in existing buildings and bridges is a challenging task. In many structures such as bridges, parking garages, ports and piers, mines, etc. columns corrode and require repair. In other cases, column jacketing is used as a technique to strengthen existing columns by adding a shell of reinforced concrete around the column. The latter is gaining popularity in seismic upgrade of structures and the ACI Committee 369 on Seismic Repair and Rehabilitation is currently developing guidelines for such applications.

The forming of existing columns is a difficult task. The existing floor and beams above prevent the use of conventional disposable cardboard tubes because they cannot be slipped over the column. The contractor is faced with assembling a form consisting of many segments around the host column. Because these forms offer virtually no resistance to the lateral pressure of the freshly cast concrete, the segments must be tied externally with bolts, clamps and the like. These forms can become very heavy and add significant time and expense to the project. The problem is especially arduous when the columns are not easily accessible.

This paper describes a recently developed product by the author that can overcome many of the above shortcomings.

FRP LAMINATE FORM

The new product is made of Fiber Reinforced Polymer (FRP). Using a special equipment, glass fibers are impregnated with vinyl ester resin and subjected to heat and pressure to make very thin laminates. For brevity, the product will be referred to as FRP Laminate Form (FLF) in this paper. The laminates have a uniform thickness that varies from 0.045 to 0.075 in. (1.1-1.9 mm) depending on the product style.

FLF is manufactured in rolls up to 102-in. (2590 mm) wide. A typical roll may include 500 feet (152 m) of FLF (Fig. 1). The laminates weigh between 0.31 to 0.51 lb/ft2 (1.5-2.5 kg/m2 ). The lightweight allows for easy handling. The mechanical properties of the laminates and the applicable ASTM Standards are listed in Table 1. The unique design of the laminate provides a perfect balance between a smooth finish surface and enough friction to prevent sliding of the surfaces.

HOW TO INSTALL FLF

The behavior of FLF is based on principles of belt friction. Various shape and size spacers are available from QuakeWrap, Inc. as shown in Fig. 1. To repair a column, for example, these spacers can be threaded through a zip-tie and secured tightly around the column. They serve to hold the reinforcing bars in place, and to provide the desired standoff distance, i.e. the annular space, between the FLF and the exterior face of the column (Fig. 2).

Property	US Customary
Flexural strength (ASTM-D790)	1.91×104 psi
Flexural modulus (ASTM-D790)	6.0×106 psi
Tensile strength (ASTM-D638)	1.0×104 psi
Tensile modulus (ASTM-D638)	5.2×105 psi
Izod impact (ASTM-D256)	4.5 ft-lb/in. notched
Coeff. of Linear Thermal Expansion (ASTM-D696)	1.7×10-5 in./in./∘F
Water absorption / 24 hours (ASTM-D570)	0.3% @77∘F
Coefficient of Static Friction	0.18

A piece of the laminate long enough to be wrapped 2-3 times around the column is cut from the roll of laminate and is tightly wrapped around the structure. A piece of duct tape can be used to secure the interior edge of the laminate to itself to prevent any cement paste getting between the layers. The free exterior end of the laminate can also be optionally secured with a few pieces of duct tape. Alternatively, a few pieces of string can be tied around the tube (Fig 3a) to maintain the size of the tube. Note that the duct tape or strings are not required to resist any loads from the internal pressure of the concrete.

Once the FLF is in place, the concrete can be placed either using a hose and tremie method or by pumping it using the optional grout ports shown in Fig. 1. The latter will create a hole in the laminate that must be patched later and will ultimately limit the number of times the laminate can be used. After the concrete hardens, the strings are untied, and the laminate is removed (Fig. 3b). The laminate can be cleaned and washed (Fig. 3c) before it is coiled and stored away for future use to build a form of the same or different shape and size. Note the finished surface of the cast concrete that is very smooth (Fig. 3d) and free from unsightly spiral marks that are commonly left behind when cardboard tubes are used as forms.

As stated earlier, the behavior of this product is unique in that instead of using external clamps to hold the form together, the layers get pressed against one another due to the internal pressure of the grout. Referring to Fig. 4, if a tube of height H is filled with concrete, having a density , an internal hydrostatic pressure p=  H is developed at the base of the form. This internal pressure shown with red arrows in Fig. 4, results in a friction force between all layers of the laminate that is shown with smaller blue arrows in that figure. The higher the pressure of the grout, the larger the friction force between the layers of the laminate. If the laminate is wrapped 3 times around itself, the external ties carry zero load. It is possible to use fewer layers. But in that case, stronger external ties such as ratchet straps must be used to resist some of the hydrostatic pressure of the concrete.

The laminates themselves are very strong and for most applications a single layer can resist the hoop stresses generated in the form. For example, assuming an 8-ft (2.4 m) high concrete placement with a unit weight of 145 lb/ft3 (23 kN/m3 ), the hydrostatic pressure at the base of the form will be 1160 psf (55.5 kPa). Assuming the diameter of the form is D=36 in. (914 mm), the tension force in a single layer of laminate will be T=145 lb (645 N). The 0.060in. (1.5 mm) thick laminate shown in Fig. 3, has a breaking strength of 600 lb/in. (105 N/mm) width of laminate which is significantly larger than the force T calculated above. This demonstrates that the construction of FLF with 2 or 3 layers of laminate is primarily aimed at enhancing the rigidity and stiffness of the tube since the tensile strength of a single layer of laminate is much more than what is needed to confine the hydrostatic pressure of the freshly placed concrete.

In the above demonstration, a circular column was formed around a damaged circular column. A major advantage of FLF is that by changing the size of the spacers, the flexible laminates allow construction of virtually any shape and size form in the field. Figure 4 shows a rectangular column cross section that can be enlarged to a larger rectangular column or an oval shape column, for example. The only limitation is that the radius of bend at the corners of a rectangular form must be larger than the allowable limit for the laminate. This radius is a function of the thickness of the laminate and is typically around 1 in. (25 mm).

PRESSURE TEST

To determine the adequacy of this system in resisting internal pressures, the following test was conducted. A 9.5- ft (2880 mm) long piece of a 0.06 in. (1.5 mm) thick laminate was cut from a long roll. This laminate was coiled to create a 3-ply tube with a diameter of 12 in. (305 mm) and a length of 102 in. (2590 mm). The free end of the laminate was secured to itself with a few short pieces of duct tape to make sure the tube diameter could not change freely.

An elongated balloon that is frequently used in internal repair of pipes was inserted inside the FLF. The balloon was connected to an air hose and gradually inflated to a pressure of 3300 psf (158 kPa). This is nearly 3 times higher than the pressure calculated in the above example and it is significantly higher than any anticipated pressure encountered in the field. The test was stopped at that pressure to avoid damage to the balloon. There was no sign of any damage in the FLF at the conclusion of the test.

SUSTAINABILITY

The FLF presented here has several unique features that contribute to making it an environmentally friendly product. One feature is the size of the FLF. With a single roll of laminate forms of any size can be built on site. The other is the adjustability of the shape of the form. As demonstrated in Fig. 4, the form can be shaped in rectangular, circular and an endless range of geometries in between. This flexibility can add significant value to many projects. Both of the above features contribute greatly to reduced transportation and storage space demand, cutting down on the number of trips to the supply stores.

The light weight of FLF eliminates the need for heavy lifting equipment on site that may be required for bulkier steel or timber forms. FLF is also fully water resistant, making its storage easier and allowing for challenging forming of submerged columns and piles. The smooth finish of FLF leaves no unsightly marks behind and eliminates the need for grinding of the finished surface.

The most significant advantage of FLF is the fact that the laminates can be used tens of times, producing forms of various shapes and sizes. The laminates have two identical sides or faces. On each side, there are two ends that can be put in contact with the concrete being placed. The only damage to the laminates can come from scratching where it meets concrete, Thus, if one end can be used 15 times, the entire laminate can be used 4 times as many or 60 uses!

Although this paper has focused on repair of existing columns, FLF can be used to build new columns as well. In those cases, an exterior frame may be used to help maintain the desired geometry of the column being cast (Fig. 5).

SUMMARY

A new type of form called FRP Laminate Form (FLF) has been presented. The form can be used for repair of existing columns or to form new columns. Among the advantages of the form are its ability to be used many times to create forms of virtually any shape and size. FLF reduces storage space and transportation cost significantly. It is fully water and rain resistant. All these features result in an environmentally sustainable product.

ACKNOWLEDGEMENT

The products and system presented in this paper are subject of multiple US Patents by the author. A video showing this product is available at: www.tinyurl.com/MoTubes

DOWNLOAD PDF

Reinforced concrete columns in many older buildings may require strengthening. This need could arise from a variety of conditions. In warm and humid coastal regions and aggressive environments, the corrosion of reinforcing steel results in the loss of capacity of the columns. In other cases, poor quality control during the original construction may result in low compressive strength of the concrete and reduced capacity of the column. The author was personally involved with the retrofit of two such buildings in Florida, USA, where the concrete compressive strength was below 1500 psi (10.3 MPa), only a fraction of the strength specified in the design documents. Some of the investigations following the collapse of Champlain Towers in Surfside, FL, have also mentioned weak and “powder-like” concrete in the columns as a potential contributing factor to the failure. This article describes a method for enhancing both the axial and flexural capacity of such columns. Implementing the technique is relatively easy, leading to a fast and economical solution with minimal disruption to the occupants. An additional feature of the repair is its small footprint, which minimizes floor space loss due to such modifications.

Conventional Solution

The author introduced the concept of repair and strengthening of structures with fiber-reinforced polymer (FRP) products in the late 1980s.1 In that approach, known as a wet layup, sheets of carbon or glass fabric are saturated in the field with epoxy. They are bonded to the external surface of structural elements, such as beams, columns, or walls. Within several hours, the materials harden and the FRP strengthening serves as additional tension reinforcement that can contribute to the flexural and shear resistance of the host structure. The FRP fabrics wrapped around the column confine the concrete and can increase its compressive strength. This results in an increase in the axial capacity of the column. While the technique is efficient for circular columns, the gain in axial capacity for rectangular columns is limited. Because FRP cannot be easily extended through the floors, it is difficult to achieve significant axial and flexural enhancement of columns with these products. Furthermore, externally bonded FRP does not increase the stiffness of the column that much. These shortcomings can be overcome using FRP laminates.

FRP Laminates

Over a decade ago, a new type of FRP laminate (PileMedic®, shown in Fig. 1) was introduced for applications in strengthening columns or piles and pipes.2 These laminates are produced with specially designed equipment where sheets of carbon or glass fabric up to 9 ft (2.7 m) wide are saturated with resin and passed through a press that applies uniform heat and pressure.

The laminates offer several advantages compared to the fabrics used in wet layup applications:

• Strength in both longitudinal and transverse directions, with tensile strength up to 155 ksi (1070 MPa), by using a combination of unidirectional and/or biaxial fabrics;
• Can be made as thin as 0.03 in. (0.8 mm), which allows them to be bent around a corner with a radius of 2 in. (50 mm);
• Manufactured in plants under high quality control standards, which improves the quality of the finished product;
• Strength can be tested before installation, which assures the design engineer that the specified strength is met, eliminating delays for corrective actions;
• Repairs can be completed much faster in the field;
• The number and pattern of the layers of fabrics in the laminates can be adjusted to produce an endless array of customized products that can significantly save construction time and money;
• They are used to build a structural stay-in-place form around the column, creating an annular space that can be filled with concrete and reinforcing bars3 and providing shear reinforcement and confinement for the column; and
• Specially designed spacers4 are used to hold longitudinal reinforcing bars in place and to help create a shell around the column (Fig. 2).

Application

PileMedic laminates are supplied in 4 ft (1.2 m) wide rolls to any desired length (wider rolls are also available). Typical detail requires the laminate to be wrapped two complete times plus an 8 in. (200 mm) overlap around the column. The laminate is cut to the desired length, and an epoxy paste is applied; the laminate is wrapped around the column and bonded to itself to create a two-ply shell at a distance of 1 to 2 in. (25 to 50 mm) from the face of the column. Additional 4 ft laminates are similarly installed and overlap the previous shell by 3 to 4 in. (75 to 100 mm) to cover the full height of the column (Fig. 3). Finally, the annular space between the column and the PileMedic jacket is filled with concrete or grout using a pump or the tremie method.

Lateral ties

The jackets also act as supplementary steel ties, which is a shortcoming in many older or corrosion-damaged columns. Eliminating the need for ties around the longitudinal bars is a great advantage that results in easy installation.

Corrosion protection

The system provides an impervious jacket around the column that prevents the ingress of moisture and oxygen. By depriving the column of exposure to moisture, the corrosion rate is drastically reduced, resulting in the long service life of the repair.

Joint region

The retrofit of the frame, especially in seismic regions, requires attention to the beam-column joint region as well. One option is to epoxy anchor steel ties into the core of the column to provide support against buckling for the newly installed longitudinal column bars (Fig. 4). This region can subsequently be encased in concrete with additional reinforcement. Such enlargements are typically within the depth of the beam and can remain invisible above the ceiling. An earlier study demonstrated that as the flexural strength ratio increases, the required lateral ties in the joint region may be relaxed.5 Thus, the flexural strengthening of the column may result in easier retrofit for the joint.

Lower construction cost

The PileMedic retrofit solution has many inherent advantages that lead to cost savings. For example, the entire system comprises lightweight materials that can be taken to any floor of the building using passenger elevators. Handling of the laminates to wrap them around the column requires no heavy lifting of equipment. The adjustability of the jacket size in the field leads to a smaller footprint and eliminates construction delays due to shipping the wrong size formwork to the site. The strength of the laminate that eliminates steel ties results in faster and less costly repairs. The estimated cost to retrofit a typical column is well below $10,000.

Design Examples

Two retrofit alternatives are presented here. In both cases, the corners of the column that do not include any reinforcing steel can be easily cut and removed to minimize the enlargement of the column and loss of floor space. Two new No. 8 (25 mm) bars can be placed at each corner, and these bars extend to the floor above through the slab. This increases the flexural capacity of the column to ensure a “strong- column/weak-beam” at that location. Plastic spacers are attached on the column to define the annular space.

Option 1

In this case, the 18 x 18 in. (460 x 460 mm) column is enlarged to a 21 x 21 in. (533 x 533 mm) column (Fig. 5(a)). A biaxial glass FRP laminate is used to create a two-ply shell around the column.

The minimal enlargement is sufficient to accommodate the eight new reinforcing bars being installed. The shell around the column is made with two plies of PileMedic glass laminate, which represents the minimum number of layers for such applications.

The interaction diagram for the retrofitted column is calculated and is shown in the solid red line in Fig. 6, assuming the grout strength is 4000 psi (27.6 MPa). The axial capacity of the column is increased by 51% from 1460 to 2215 kip (from 6500 to 9850 kN). The flexural capacity is also increased by 220% from 215 to 485 kip∙ft (from 291 to 657 kN∙m). Therefore, the flexural strength ratio for the retrofitted frame is

This is larger than the minimum value of 1.2 and ensures that any plastic deformations are concentrated at the beam ends.

Option 2

If, in addition to flexural capacity enhancement, a greater increase in the axial capacity of the column is also desired, it’s best to alter the column into a circular section (Fig. 5(b)). Because confinement is a function of the stiffness of the jacket, a carbon laminate can be used instead of the glass laminate used previously. A circle with a diameter of 23.7 in. (600 mm) has the same area as the 21 x 21 in. column used in Option 1, that is, the footprint of the repair for both options is the same. However, the combination of circular geometry and wrapping with the stiffer and stronger carbon laminate results in an increase in the compressive strength of the original concrete and the newly placed concrete in the annular space. ACI 440R.2-176 provides guidelines for quantifying this strength gain and, for this example, the confined concrete reaches a compressive strength of 5150 psi (35.5 MPa).

Retrofit	Option 1	Option 2
Laminate type	PLG14.13	PLC150.10
Laminate construction	Biaxial glass	Unidirectional carbon
Tensile strength, ksi	28.7	156
Tensile modulus, ksi	2840	13,800
No. of plies in wrap	2	2
Equivalent lateral tie	No. 4 Grade 40 at 3.7 in.	No. 4 Grade 40 at 1in.
Original column f”c, psi	4000	4000
Enlarged shape	21 x 21 in. (square)	23.7 in. (round)
Enlarged area, in.2	441	441
Confined f”cc, psi	4000	5150
Pn, kip	2215	2633
Mn, kip-ft	485	499

The compressive strength of concrete does not affect the flexural strength of the column significantly. In this case, the retrofitted column has a flexural capacity of Mn = 499 kip∙ft (676 kN∙m), which is slightly higher than in Option 1. However, as shown in the interaction diagram (Fig. 6), the axial capacity of the confined column increases greatly. In this case, an 80% increase from the original column and a 19% increase, when compared to Option 1 with a square shell with glass laminate, is achieved. Clearly, this option is preferred when the gain in axial capacity of the column is also desired. For example, this could be the preferred retrofit method when due to construction errors, the compressive strength in the column is lower than the specified value.

A summary of these retrofit alternatives is presented in Table 1.

Lateral Tiles

ACI 440.2R-17 provides environmental reduction factors for FRP based on the use conditions, such as exterior versus interior installation and the type of fibers used, carbon versus glass. Including these reduction factors, the glass laminate is equivalent to providing No. 4 Grade 40 (275 MPa) ties at a spacing of 3.7 in. (94 mm), while the carbon laminate is equivalent to No. 4 ties at a spacing of 1.0 in. (25 mm); refer to Table 1. In both cases, these values exceed what the current codes require.

Footprint

The footprint of the proposed retrofit is very small. In this example, the column cross-sectional area was increased by 36% for both the square and circular alternatives, while the flexural capacity of the column was more than doubled.

Field Application

Since the introduction of this system, many agencies have conducted independent tests to verify the efficacy of these laminates for a range of applications. These include a study funded by the National Science Foundation (NSF) and the California Department of Transportation (Caltrans) for fast repair of earthquake-damaged bridge piers, a study funded by the Nebraska Department of Roads for strengthening deteriorated timber bridge piles, and another funded by the Texas Department of Transportation for the repair of corrosion-damaged steel H-piles. The most significant investigation was a 3-year study by the U.S. Army Corps of Engineers, which resulted in the military selecting a laminated product to repair submerged piles worldwide. The U.S. Navy’s website reported that the product was used to repair concrete piles in Ukraine (www.tinyurl.com/PLM-UKR). The U.S. Army Corps of Engineers and the Federal Emergency Management Agency (FEMA) have also singled out these laminates as the selected product for repairing columns and piles that may be damaged in a disaster, including hurricane, earthquake, terrorism, and more in its 2013 Field Operations Guide.

References

1. Saadatmanesh, H., and Ehsani, M.R., “Fiber Composite Plates Can Strengthen Beams,” Concrete International, V. 12, No. 3, Mar. 1990, pp. 65-71.
2. Ehsani, M.R., “FRP Super Laminates,” Concrete International, V. 32, No. 3, Mar. 2010, pp. 49-53.
3. Ehsani, M.R., “Reinforcement and Repair of Structural Columns,” U.S. Patent No. US 9,890,546 B2, Feb. 13, 2018, 13pp.
4. Ehsani, M.R., “Spacers for Repair of Columns and Piles,” U.S. Patent No. US 10,808,412 B2, Oct. 20, 2020, 18pp.
5. Ehsani, M.R., and Wight, J.K., “Confinement Steel Requirements for Connections in Ductile Frames,” Journal of Structural Engineering, ASCE, V. 116, No. 3, Mar. 1990, pp. 751-767.
6. ACI Committee 440, “Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures (ACI 440.2R-17),” American Concrete Institute, Farmington Hills, MI, 2017, 112 pp.
7. Yang, Y.; Sneed, L.; Saiidi Saiidi, M.; Belarbi, A.; Ehsani, M.; and He, R., “Emergency Repair of an RC Bridge Column with Fractured Bars using Externally Bonded Prefabricated Thin CFRP Laminates and CFRP Strips,” Composite Structures, V. 133, Dec. 2015, pp. 727-738.
8. Gull, J.H.; Mohammadi, A.; Taghinezhad, R.; and Azizinamini, A., “Experimental Evaluation of Repair Options for Timber Piles,” Transportation Research Record, V. 2481, No. 1, Jan. 2015, pp.124-131
9. Dawood, M.; Karagah, H.; Shi, C.; Belarbi, A.; Vipulanandan., C.; Bae, S.-W.; and Lee, S., “Repair Systems for Deteriorated Bridge Piles: Final Report,” FHWA/TX-17/0-6731-1,Texas DOT, Austin, TX, , Apr. 1, 2017, 538 pp.
10. Hammons, M.I; Strickler, J.S.; Murphy, J.W.; Rabalais, C.P.; Crane, C.K.; and Barela, C., “Pile Wrapping for Expedient Port Repair,” Draft Report, U.S. Army Corps of Engineers, Vicksburg, MS, Aug. 2018, 117 pp
11. “Field Operations Guide,” seventh edition, U.S. Army Corps of Engineers, Vicksburg, MS, June 2013, pp. 4-4 to 4-5.

DOWNLOAD PDF

ABSTRACT

Steel frames utilizing Buckling Restrained Brace (BRB) elements have become popular in the last two decades. However, few practical solutions have been offered to retrofit braces in the large inventory of existing steel frames. These vulnerable non-seismic conforming concentrically braced frames pose a significant risk and challenge. This paper presents a new technique for retrofit of concentric braced frames using a recently-developed type of Fiber Reinforced Polymer (FRP) laminate. The lightweight laminates require no heavy equipment for installation, making the retrofit ideal for buildings with limited access. Results from preliminary large scale physical tests of the retrofit scheme are presented. The materials cost for the retrofit were less than $1000. The tests indicate the potential for a cost-effective, and easy-to-implement retrofit solution for the large inventory of seismically- vulnerable existing steel buildings

Introduction

Concentrically-braced frames (CBFs) comprise a significant portion of the national building inventory. The popularity of the CBF as a Lateral Force Resisting System (LFRS) stems from its economy through its high stiffness, requiring placement of braces in only a few bays of the overall floor plan at each floor. These bays can often be located along a stairwell, bathroom or elevator core, rendering the remainder of the floor plan open architecturally and designed economically for gravity load alone. CBFs are configured as single diagonals, chevrons, and single or two-story X-braces. Common CBF bracing elements for moderate forces include bolted double angles members that can “knife-fit the gusset, welded channels, tees, and in-cases, tension-only straps. For heavier loads, tubes or wide flange sections are used.

Steel CBFs are used extensively in buildings as a LFRS due to their high elastic stiffness and structural efficiency. However, the seismic performance of CBFs, which is dominated by the cyclic inelastic behavior of bracing members, is less robust than other common steel LFRS. The CBF brace elements act in both axial compression and tension, due to the cyclic nature of lateral forces (wind, earthquake), and thus the CBF brace element is susceptible to buckling in the earthquake.

Much research has been devoted to achieving higher levels of seismic performance in CBFs. As a result, modern braced-frame structures in regions of high seismic hazard are designed as Special Concentrically-Braced frames (SCBFs). SCBFs are designed for lower forces than required to remain elastic in the earthquake, and thus count on special detailing: compact sections for the brace member (e.g. tubes) to minimize the detrimental effects of instability, and carefully detailed connections to avoid fracture under tension. More recently, a high seismic solution gaining popularity is the Buckling Restrained Brace (BRB), which prevents buckling altogether.

Nevertheless, thousands of older non-compliant CBF structures still exist in regions of high seismic hazard, built in an era when seismic effects were not well understood, and thus prior to the creation of the stringent SCBF requirements. Further, in regions of low to moderate seismic hazard, in addition to older structures, most modern CBFs are designed without special seismic details, i.e. as Ordinary CBFs (OCBFs). These structures do not have the expected performance to survive a rare but destructive earthquake [1]. With the severe societal costs of rare but destructive Maximum Considered Earthquake (MCE) events [2], including the recent recognition of the increased risk for those events in regions of low to moderate seismicity [3], these structures represent a significant risk to the economy of the U.S. and safety of the general public.

The reason these vulnerable CBFs remain un-retrofitted in the U.S. building inventory, is due to the difficulty in achieving an economical yet effective retrofit for these structures. While retrofit solutions to provide ductile detailing is straightforward conceptually, the cost premium and practical limitations renders most retrofit schemes unworkable. In short, there are currently no both economical and effective retrofit techniques for low ductility CBFs [1]. This paper presents a recently-developed technique that allows economical retrofit of CBFs and some the preliminary test results.

Buckling-Restrained Braces (BRBs)

The BRB is a seismic bracing element that responds in ductile fashion to an earthquake. BRBs use an unbonded confining tube around a slender steel bracing core (Fig. 1a), thus allowing the core to yield in tension, but suppressing buckling in compression [4]. The typical BRB is: (1) a single piece yielding steel core with stiffer non-yielding end sections; (2) a restraining mechanism, often a steel tube with filled concrete with an end gap to permit permanent elongation. Under tension, the BRB behaves as a traditional steel brace; however, under compression, the yielding core is prevented from buckling by the restraining element, resulting in ductile, symmetric load behavior with high energy dissipation and compression strength than a CBF (See comparison in Fig. 1b). The BRB has been widely studied and is now commonly used for new construction of special seismic frames. The high-performing BRB is quite expensive ($25K/unit) [4].

BraceWrap® Retrofit Concept

The BraceWrap® retrofit concept is a recently patented system [5] that converts an existing nonductile steel brace into a buckling-restrained brace. The technique is made possible by a special type of thin Fiber Reinforced Polymer (FRP) laminates developed a few years ago by one of the authors [6]. The laminates are constructed with specially-designed equipment. Sheets of carbon or glass fabric up to 60-in. wide are saturated with resin and passed through a press that applies uniform heat and pressure to produce the laminate (Fig. 2). Depending on the type of fabric (glass or carbon), the laminates provide a tensile strength ranging from 60 to 155 ksi. Typical thickness of the laminates is 0.025 in., resulting in a flexible laminate that can be bent by hand (Fig. 2).

Since their introduction, the versatility of these laminates has led to be used in many structural rehabilitation applications worldwide, including bridge piers [7], submerged piles [8], cell phone towers [9], and pressure pipes in gas industry [10]. The thin laminates are flexible enough that can be wrapped around a brace to create a tube as small as 6-in. in diameter. A roll containing 200-feet of the laminates can be easily carried to anywhere in the building where the retrofit is required.

Retrofit Procedure

Due to space limitation, the retrofit steps are shown using the photos from a specimen that was tested as described in the following section. To ensure proper behavior as a BRB, a debonding agent should be used to prevent composite action between the brace and the concrete. In this case, duct tape was applied to the brace and it was coated with a bond-breaking material. Special plastic spacers were used along the edges of the brace to make sure the laminate does not come in contact with the brace (Fig. 3a). The 4-ft wide laminate is cut in a length equal to twice the perimeter of the tube to be made, plus 8 inches. An epoxy paste is mixed and applied to the laminate except the portion that will be directly facing the brace (Fig. 3b). The laminate is wrapped around the brace and bonded to itself to create a two-ply tube of desired diameter.

Because the epoxy has not cured yet, the size of the tube can be adjusted and temporarily held in the desired size using a few ratchet straps that are wrapped around the tube. The bottom of the tube is sealed, and the tube is filled with concrete. The next laminate is similarly installed, with an adequate overlap length that is epoxied to the first tube. This process is continued until the entire brace is retrofitted in 4-ft increments.

The technique offers several unique advantages:

• It is applicable to both retrofit of existing frames and construction of new structures; the latter will reduce the cost of BRBs in new buildings that is currently controlled by few suppliers.
• The low-cost solution is very easy to implement in the field.
• Repairs require no major equipment and can be performed with a 2-man crew.
• Unique design is one-size-fits all, and eliminates costly custom-made products.
• Materials are lightweight and can be readily delivered to upper floors of a building using passenger elevators and stairways.
• Building safety is not compromised since the existing braces are not removed during the modifications.

Pilot Testing for Demonstration of Concept

A pilot test for BraceWrap proof of concept performed at the UA Laboratory. The test objective was to directly compare the performance of a CBF w/ and w/out the retrofit. The brace element was a welded double angle LL3½x3½x¼ with effective slenderness KL/r = 87.9. The BraceWrap retrofit (See Fig. 4c) used a 60ksi glass laminate formed into an 8¼” cylinder shell, held in place by plastic spacers (3D printed in-house), and filled with high-strength non-shrink quickset grout. The plies were held together using in-house epoxy paste. Surface preparation used household products: The ½” separation between angles was first filled with grout. The bond breaker was duct tape smeared with high temperature wax release. The grout, held in with a thin Polythene sheet, hardened in 3hr; the epoxy cured in 12hr. The total installation took two workers less than 3 hours. The materials cost for the BraceWrap® laminates, the epoxy paste, and the grout was less than $1000.

The unretrofitted brace incurred flexural-torsional buckling, leading to a rapid drop of load, followed by local buckling of the plastic hinge region, and after a few bucklingstraightening cycles, a low-cycle fatigue fracture through the entire outstanding flange of the hinge region (Fig. 4a). BraceWrap® successfully increased the buckling load of the brace by over 40%, significantly increased energy dissipation, and suppressed the local buckling and lowcycle fatigue failure of the brace (Fig. 4b). Just this improvement can make the difference between significant damage and modest damage in a CBF structure, or even prevent collapse for a highly vulnerable CBF.

However, lessons learned in the test provided insight into oversights in the installation process for the pilot test that are easily remedied, and should lead to significant further increases in strength and performance. Foremost was the inattention to the location of joints between the laminate sheets. For convenience, the sheets were placed from the bottom of the brace (with grout inserted in stages). As such, one of the joints was located within 2½” of the brace midspan. This joint failed (Fig. 4d), leading to the loss of load carrying capacity. It is believed that simply centering one of the sheets to straddle the mid-span would be enough to eliminate this failure. However, increasing the minimal laps (12” overlap) and epoxy used in this first trial are also parameters that warrant further investigation. It was clear from the test that the single sheet of laminate itself is strong enough to carry the hoop stresses; however, a second layer of staggered sheets over the joints is always possible at minimum extra cost. Finally, the desire to minimize the cross-section left the minimum grout cover 3/8” at the outstanding flanges. While it is not believed this led to the failure, the 8¼” diameter could easily be increased to provide more cover, and will be a design parameter.

It is noted that the bond breaker scheme fulfilled its objective, as the elastic stiffness of the retrofitted frame was similar to the original frame, and the core slipped +0.12/-0.09” on the brace. An additional identical specimen is currently in preparation where the above issues will be addressed. It is believed that by removing the overlap joints away from the midspan of the brace and/or increasing the number of wraps from 2 to 3 significant enhancement in the behavior of the retrofitted system will be observed. These modifications will have an insignificant impact on the cost of the proposed retrofit system.

Conclusions

This paper presents the undergoing development of a new cost-effective solution to convert an ordinary concentrically braced frame to a buckling restrained braced frame. The results of a pilot experimental study have been presented. Two welded double angle braces were tested, one as a control specimen and the other as a retrofitted frame. The experimental results indicate an increase of greater than 40% in the retrofitted specimen compared to the control specimen. However, the retrofitted specimen failed prematurely due to a design error by placing an overlapping joint at the midspan of the brace. An additional test is underway where the joint will be removed away from the midspan. It is anticipated that the performance of that specimen will be significantly improved compared to the first test.

The proposed retrofit system could be accomplished in 3 hours with a two-man crew and the materials cost were below $1000. The installation requires no heavy lifting equipment and can be easily carried out in locations where access is limited. These factors indicate that the propose system could have a significant value on reducing the seismic vulnerability and enhancing the safety of many existing structures.

References

1. Fahnestock L, Hines E, Tremblay R, et. al.. Reserve Capacity & Implications for Seismic Collapse Prevention for Low-Ductility Braced Frames in Moderate Seismic Regions. Proceedings of the 10th NCEE, Alaska, 2014.
2. Fleischman R, Restrepo J, Pampanin S, Maffei J, K.Seeber K, and Zahn F. Damage Evaluations of Precast Structures in 2010-11 Canterbury Earthquake. Earthquake Spectra, 2014; 30(1):277-306.
3. Peterson M, Moschetti M, Powers P, Mueller C, Luco N, et.al.. Documentation for the 2014 Update of the United States National Seismic Hazard Maps. USGS, 2014.
4. Star Seismic. 6300 N Sagewood Dr, Park City, UT 84098. https://www.linkedin.com/company/star-seismic/about/
5. Ehsani, M. 2017. Buckling Reinforcement for Structural Members, US Patent No. 9,719,255, U.S. Patent and Trademark Office, Alexandra, VA, 15 pp.
6. Ehsani, M. FRP Super Laminates: Transforming the Future of Repair and Retrofit with FRP. Concrete International 2010; ACI, 32(03): 49-53.
7. Ehsani, M., and Croarkin M. A Novel Solution for Restoration of Deteriorated Piles. Government Engineering, March-April 2011, 14-15.
8. Ehsani M, Day S, and White T. Repair of ASR-Damaged Piles in a Crocodile-Infested River.” Concrete International 2017; American Concrete Institute, Detroit, MI, 39(4): 61-65.
9. Ehsani M, and Rimmele RJ. Structural Challenge Facing the Wireless Communications Industry – A New FRP Solution for Strengthening Concrete Telecommunication Towers. Structure Magazine 2016; 23(12):22-24
10. Carbone M, Ehsani M, and Ragula G. Winter Wonderland Does Wonders to CIPP Renewal of a High-Pressure Gas Main in New Jersey, North American Society for Trenchless Technology (NASTT) No Dig Show 2012; Nashville, TN, Paper E-1-01, 10 pp.

DOWNLOAD PDF

Case Study

A 55-ft (16.8 m) tall hollow precast concrete cell phone pole located in Los Angeles (Fig.I) was originally designed and constructed in 1996. The prestressed concrete spun-cast pole had a constant outside diameter of 15.75 in (400 mm) with a wall thickness of 2¼ in (57 mm). Reinforcing steel is comprised of0.375 in (10 mm) diameter 250k (1112 kN) seven-wire strands. The structure is supported on a 36 in (914 mm) drilled pier foundation embedded 17 ft (5.2 m) deep.

Based on the code requirements at the time of design, prestressed concrete poles were typically designed for a single carrier level. Future proposed antennae additions coupled with code changes since the design can lead to the structure requiring additional capacity. Options for reinforcing these structures are limited and typically not cost fea­sible. The alternatives that were considered for this project are reviewed below.

The original capacity of the pole was 1135 kip-in and remains constant throughout the entire height of the pole (Fig. 2). The pole was analyzed for all dead and live load effects, including wind and earthquake . The control­ ling new demand shown in Fig. 2 resulted in the lower 22 ft (6.7 m) of the pole being overstressed by various degrees . As such, the strength of the pole needed to be enhanced by nearly 136% at the base.

One option used on past projects requires guying the structure that presents technical hurdles as well as practical considerations. The introduction of guy wires will most likely require the tower owner to incur additional costs purchasing or leasing additional ground space in the vicinity of the tower in order to expand the site foot print to place the new guy anchorage points. The guy wires would also require additional ongoing maintenance and inspections.

A second option has been to build a steel tower, including new foundations , around the existing structure. This option is typically quite expensive and may be prohibited by the local permitting jurisdiction due to significant changes to the aesthetics of the original structure.

In some cases, a third option to remove and replace the structure may be considered. However, this alternative is typically not preferred due to the logistics, the required permitting process, cost considerations and the disrup­tion of service to the wireless providers. Complicating all of these options is that typically there is minimal space to work within the existing congested site compound with numerous obstructions.

The fourth option involved the use of Fiber Reinforced Polymer (FRP) that is described in more detail below.

The original capacity of the pole was 1135 kip-in. and remains constant throughout the entire height of the pole (Fig. 2).  The pole was analyzed for all dead and live load effects, including wind and earthquake.  The controlling new demand is shown in Fig. 2.  The lower 22 feet of the pole were overstressed by various degrees. At the base, the strength of the pole had to be enhanced by nearly 136%.

One option used on past projects requires guying the structure that presents technical hurdles as well as practical considerations.  The introduction of guy wires will most likely require the tower owner to incur additional costs purchasing or leasing additional ground space in the vicinity of the tower in order to expand the site foot print to place the new guy anchorage points.  The guy wires would also require additional ongoing maintenance and inspections.

A second option has been to build a steel tower, including new foundations, around the existing structure.  This option is typically quite expensive and may be prohibited by the local permitting jurisdiction due to significant changes to the aesthetics of the original structure.

In some cases a third option to remove and replace the structure may be considered.  However, this alternative is typically not preferred due to the logistics, the required permitting process, cost considerations and the disruption of service to the wireless providers. Complicating all these options is the fact that typically there is minimal space to work within the existing congested site compound with numerous obstructions. The fourth option involved the use of FRP that is described in more detail below.

FRP Alternative

Among the options considered, the use of FRP offered the most viable solu­tion (Fig. 3). The technique offered the flexibility to readily change the strength of the pole along the height. A close examination of the stresses revealed that both compressive and tensile stresses in the pole exceeded the allowable limits under the new loads. FRP products have high tensile strength and can address the shortcomings of the pole easily. However, the compres­sive strength of FRP is significantly lower than its tensile strength. In most retrofit projects, it is not uncommon to ignore the compressive strength of the FRP. This meant that the thickness of the concrete wall of the pole had to be increased by the addition of conventional concrete or grout to lower the compressive stresses below allowable limits.

Tension reinforcement for the pole was provided by bonding unidirectional carbon fabrics to the exterior surface of the pole. The carbon fabric used for this project was supplied in 24 in (610 mm) wide rolls and has a tensile strength of over 6 kips (27 kN) per 1 inch (25 mm) width of fabric. The fabric could also be cut into narrower bands for ease of installation without any adverse effect on its strength.

Based on the required strength , three (3) layers of carbon fabric saturated with epoxy were applied on the lower 16 ft (4.9 m) of the pole. Two (2) layers were sufficient for elevation 16 to 26 ft (4.9 to 7.9 m). Elevations above 26 ft (7.9 m) required no strengthening. For confinement and to eliminate any interference with the antennae signals, 2 ft (0.6 m) wide bands of a unidirectional glass fabric saturated with epoxy were wrapped in the hoop direction over all carbon fabric (Fig. 4) for elevations O to 26 ft (7.9 m).

As mentioned earlier, the 2¼ in (57 mm) thick wall of the pole was over­ stressed in compression. Therefore, the pole wall thickness had to be increased to meet the allowable compressive stresses. This increase in thickness can be different along the height of the pole, requiring a form­ work or jacket whose diameter could be easily adjusted in the field. For this project, an increase in thickness of 1½ in (38 mm) for heights Oto 22 ft (6.7 m) was sufficient. For elevations of 22 to 26 ft (6.7 to 7.9 m), no increase in thickness was required and only FRP fabric was sufficient for increased tensile strength.

A special FRP laminate offered an economical solution for this applica­tion. The laminates are manufactured in the plant by saturating rolls of fabric with resin and running them through a special press that applies heat and pressure, resulting in 4 ft (1.2 m) wide rolls with a thickness as little as 0.01 in (0.25 mm). The relatively large width and small thickness of the laminates makes their manufacturing unique and challenging. The thin laminates (Fig. 5) are flexible and could be readily wrapped around the pole in the field to create a stay-in-place form that can be filled with grout or resin. The properties of the laminates are listed in Table 1. Depending on their composition, the laminates offer reinforcement in one or two directions.

	Unidirectional Carbon	Biaxial Carbon	Biaxial Glass I	Biaxial Glass II
Thickness, in (mm)	0.026 (0.66)	0.026 (0.66)	0.026 (0.66)	0.010 (0.25)
Longitudinal Direction:
Tensile Strength, ksi	156	101	62	49
Tensile Modulus, ksi	13,800	7,150	3,500	3,200
Traverse Direction:
Tensile Strength, ksi	9	64	60	49
Tensile Modulus, ksi	1,190	2,940	3,650	3,200

For this project, the biaxial glass laminate with a thickness of 0.026 in (0.66 mm) was used. For the region covering the height of 4 to 22 ft (1.2 to 6.7 m), these laminates were coated with an epoxy paste and wrapped around the pole to create a two-ply shell. At this stage, the shell is not bonded to the pole and it is free to be moved up and down. Temporary l½ in (38 mm) thick spacers , such as a PVC pipe, were attached to the pole surface to facilitate the wrapping of the laminates with the necessary annular space. The structural shell created in this fashion provides the equivalent of No. 4 Grade 40 ties at a spacing of 2½ in (64 mm) along the pole. The shell also offers a tensile resistance similar to No. 4 Grade 40 steel reinforcing bars distributed at 2½ in (64 mm) spacing around the pole. For this project, these contributions were conservatively ignored. The selection of glass over carbon laminates was based on the electrical insulation properties of the former.

Near the base of the pole, the moments had to be transferred into the footing. Since the carbon FRP is terminated at this location, steel rein­forcement was used to achieve this objective. The FRP fabric in that region was coated with a layer of sand for improved bond and to transfer stresses. Twelve No. 8 Grade 60 bars each 14 ft (4.3 m) long were epoxy anchored into the existing foundation. These bars extended 4 ft (1.2 m) above grade (Fig. 6). Even though a 4 in (102 mm) annular space was sufficient, the existing square steel base plates required the shell to have a diameter of 25.75 in (654 mm) at the base. As can be seen in Fig. 2, this resulted in a very conservative over design near the base of the pole. The completed repair of the base region was just wide enough to fit in the available space between the existing cables and the fence wall (Fig. 6).

Field Installation

The strengthening solution presented above took approximately four days to be completed. The lightweight laminates eliminated the need for heavy equipment and all work was accomplished using a manlift (Fig. 1) or scaffolding (Fig. 7). Once the shells were created around the pole, the annular space was filled with a high-strength non-shrink grout. The laminates were coated with a UV-resis­tant coating. Many cables and appurtenances that were near or attached to the pole were moved slightly to accommodate the FRP and concrete placement, and were then relocated to their original position (Fig. 6). The design allowed the pole to remain fully operational during the repair with little change in the appearance and size of the pole.

Fifty-two (52) monopoles have been retrofitted with this technique in 2015 and 2016. Additional structures have been scheduled for retrofit in the greater Los Angeles basin and elsewhere for 2017 based on cell phone coverage needs.

DOWNLOAD PDF

IBC 18-10

KEYWORDS: Carbon Fiber Reinforced Polymer (CFRP); Repair; Strengthening; AASHTO girder; Damage

ABSTRACT: In its first application of Carbon FRP, ADOT successfully completed the repair of 12 AASHTO Type II girders that were damaged by truck collision. The loss of flexural strength in the girders varied from 20% to 50%. The positive outcome of this project will allow ADOT to include this repair option in future similar applications.

INTRODUCTION

This paper presents the repair of two bridges that were recently completed in Phoenix, Arizona. The bridges were along interstate 1-17 which is a major freeway passing through Phoenix. The project locations are shown in Fig. 1.

The first was an underpass, 1-17 at Jefferson St. The second was an overpass, 1-17 at 19th Ave. The bridges are located in downtown Phoenix Arizona and are less than 2 miles apart. During 2010, 1-17 in this area of Phoenix had only 3 lanes in each direction with an Annual Average Daily Traffic (AADT) volume of 100,000 with trucks comprising approximately 9.6% of that volume. Traffic disruptions in the bridge vicinity have always been a major concern.

The bridge at Jefferson Street is a two simple span precast prestressed AASHTO Type II concrete girder bridge built in 1959. The bridge is 132′-9″ long composed of two 65′-4 ½” spans with an out to out width of 64 feet. The existing northwest girder over 1-17 southbound has been repaired multiple times due to vehicle collisions in 2003, 2006, 2008 and 2014 (Fig. 2). Each time the girders were repaired and the girder shape reformed to the original shape. Several of the remaining girders have sustained minor damage consisting primarily of spalls, exposed rebar and minor cracking. Only the single northwest girder over 1-17 that was severely compromised was of concern at this time.

The bridge at 19th Ave was originally two two simple span precast prestressed AASHTO Type II girder bridges built in 1961. In 1984 the median between the two bridges was closed with four precast prestressed Bl-36 box beams to form a single structure. The current bridge is 103′-9″ long composed of two 49′-3″ spans with an out to out width varying from 122′-5″ to 125′-7″. The 19th Ave bridge has endured numerous vehicle collisions. Most of these collisions are minor collisions leading to spalling at the bottom flange of the girders. Several girders have had exposed rebar and few collisions have led to exposed and damaged prestressing strands.

Due to the vast numbers of collisions and the fact that it is very unusual to catch a vehicle that has impacted the bridge, it is impossible to determine exactly how many times this bridge has been hit. Moreover, it is difficult to determine if the damage was sustained from a single collision or multiple bridge collisions. Examples of several damaged girders can be seen in Fig. 3. Although the number of collisions to girders is greater at 19th Ave than at Jefferson Street, the amount of damage to any single girder is far less.

REPAIR ALTERNATIVES CONSIDERED

The Arizona Department of Transportation (ADOT) Bridge Group believed that with these collisions had most-likely resulted in a loss of strength in many of these girders. The determination of which girders needed to be strengthened was primarily based upon previous bridge inspections and records indicating which girders had prestressing strands damaged or exposed from a previous collision. Bridge Group believed any girders that did not have prestressing strands exposed did not have any significant loss in strength and could be repaired with concrete patching to protect from loss due to corrosion.

ADOT Bridge Group needed a way to restore the damaged girders to full strength without significant disruption to traffic in the area. Because of the heavy traffic volumes, girder replacement was not a viable option. 1-17 cannot be closed that long. Fiber Reinforced Polymer (FRP) was chosen primarily because it could be installed quickly and with minimal distribution to traffic. With the possibility of a girder replacement removed, FRP seemed to be the only choice.

Strengthening of structures with FRP is a technique that was pioneered by the author in the late 1980s (Ehsani and Saadatmanesh 1990). In this technique, fabrics of arbor glass are saturated with epoxy and are externally bonded to the surface of the structural member. Within several hours, when the epoxy cures, the FRP reaches a tensile strength that is 304 times that of steel. The design guidelines by ACI and AASHTO require that the substrate be of a good quality and provide a minimum tensile strength of 200 psi. This requirement is easily satisfied for AASHTO girders that are manufactured in plants and with good quality concrete.

Each layer of FRP is typically around 0.04 to 0.05 in. thick. The fabrics are flexible during the installation phase and they conform to the shape of the girder. A further advantage of FRP is that it provides an impervious encasement for the structural member. In cases where the member is fully encased in FRP, this layer prevents the ingress of oxygen or moisture and results in substantial reduction of corrosion rate.

ADOT METHODOLOGY

Although other state DOTs had utilized FRP for strengthening of girders, the State of Arizona had never done so. An ADOT Bridge Group Engineer began the process of researching this technique could be implemented on this project. It was discovered that the City of Phoenix had performed a strengthening of bridge girders with Carbon fiber reinforced polymer (CFRP). With extensive research on the internet, contact with the bridge engineers at the City of Phoenix and discussions with CALTRANS, ADOT decided to proceed by using CFRP and had a method to perform the project.

Once the use of CFRP was selected as the method of repair, the next question was the extent of the repair for each girder. It was evident that the girders had lost strength due to impacts but without closing the bridges, attaching sensors and measuring results, an exact determination of the loss of strength was not possible. The original plan Record Drawings did not have complete strand data such as the number of strands and the strand placement information. This made it impossible to even determine the original design strength of the girders. Consequently, Bridge Group decided to base the full strength of the girder using the bridge load rating. After some calculation and using primarily engineering judgement a conservative value was assigned to each bridge on how much strength needed to be restored to bring the girders back to full strength.

The 1-17 underpass at Jefferson Street was determined to have a maximum flexural design capacity of 1988 k-ft. Since the girder at Jefferson Street had sustained multiple major impacts completely severing multiple strands Bridge Group assumed a strength loss of 50% for that girder. Thus, the FRP was supposed to provide an additional flexural capacity of 994 k-ft. It is noted that the girder on Jefferson Street was an edge girder supporting a sidewalk, not live traffic and the bridge was deemed capable of continuing to function even without that girder. At no time would ADOT risk public safety.

The 1-17 overpass at 19th Avenue Street was determined to have a maximum flexure moment design capacity of 1799 k-ft. Since the girders at 19th Avenue had only sustained minor impacts without completely severing strands, Bridge Group assumed a strength loss of 20%. Thus, the FRP solution was supposed to provide an additional flexural strength of 360 k-ft.

Bridge Group ran design calculation to verify that a CFRP design could be created that satisfied the strength requirements. Since federal money was used on this project ADOT could not choose a contractor or supplier for the work. The project had to go for an open bid. This gave Bridge Group one of its most difficult tasks, creating special provisions. Since the properties of CRFP vary between manufactures a performance based specification had to be created. This would give contractors and suppliers an equal opportunity to bid to work on the project. Because the CFRP material would vary based upon who won the bid Bridge Group could not design CFRP. Bridge Group could only call out the required performance. It was up to the contractor to provide the design that would meet the specified performance requirements.

As noted earlier, ADOT Bridge Group worked closely with CALTRANS in gathering information for Arizona’s first FRP project. Much of the work on the specifications came from specifications used by CALTRANS. ADOT Bridge Group adapted CALTANS specifications for use in Arizona, and eliminated several tests that were considered redundant. In addition, Bridge Group collaborated with the ADOT Materials Group and ADOTs internal testing labs to verify what tests Arizona wished to perform. The final decision was to secure the services of an outside independent testing firm for this project but require extra samples for the ADOT lab to test. ADOT testing labs wished to compare their results to the independent firm to verify if ADOT testing labs could perform the testing for future FRP projects.

The project was put out for an open bid. FNF Construction, Inc. was the lowest bidder and was awarded the contract. FNF Construction had chosen FRP Construction LLC (Tucson, AZ) as its FRP supplier and installer. FRP Construction LLC hired QuakeWrap Inc. (Tucson, AZ) to provide the engineering design and the FRP materials for this project.

DESIGN

The client, ADOT selected an externally bonded CFRP system for the strengthening and protection of several AASHTO Type II girders over 2 different bridges in Phoenix, AZ, 19th St and Jefferson St. QuakeWrap Inc. was selected as the CFRP engineering designer to restore the girders to their original performance capacities and develop construction documents for the installation of the repairs. ADOT, requested that moment capacity be restored to the girders and the amount of loss was up to 50% of the original capacity for the Jefferson St. bridge and 20% loss for the 19th St. bridge. Additionally, ADOT requested that 2 layers of fabric be installed similar to shear reinforcement to act as a protection system for securing the new flexural reinforcement and to act as a net holding spalled or impact damaged concrete from falling on vehicles below. QuakeWrap Inc completed all design work per the AASHTO Guide Specifications for Design of Bonded FRP Systems for Repair and Strengthening of Concrete Bridge Elements (1st Edition), as required by ADOT.

The first step in the design process required the evaluation of AASHTO Equation 1.4.4-1 which determines an element’s eligibility to be strengthened with CFRP.

After it was determined that the girders were eligible for the repair with a CFRP system, engineers began to focus on increasing the flexural strength. Section 3.4 of the AASHTO code for flexural strengthening of reinforced concrete elements was followed for design purposes. The design approach is based on moment equilibrium and thus can be amplified to be used over a variety of girder shapes. The designs used different amounts of unidirectional CFRP layers installed at the bottom and lower bottom sides of the girders to meet the client’s need and such that the factored moment resistance was higher than the factored moment demand (Fig. 4). Many assumptions were made to carry out a design. The most significant assumption that was approved by ADOT was allowing the bridge deck slab to act compositely with the girder.

The key to meeting the design requirements on the Jefferson St. heavily damaged bridge girder was the addition of the longitudinal fabric on the sides of the girder. Fabric placed solely on the typical bottom surface would not suffice in this case. Every layer of CFRP fabric added takes a significant reduction in the allowable strain of that layer and continues to reduce for every layer after the first. This is due to the fact that all layers of fabric share the bond to the substrate with the first layer that is installed. Putting fabric on the sides reduces the contribution of the fabric by reducing the moment arm from the cross-section’s neutral axis. However, having more fabric with a first layer strain contribution far outweighed the slight reduction in the moment arm. Had it not been for the side installed fabric this girder would not have been retrofitted to the requested demand.

The flexural capacity of the girders that were retrofitted were calculated with unidirectional CFRP installed in the direction of the span of the elements, thus increasing their moment capacity by providing an FRP capacity to match or exceed the missing original capacity as shown. For Jefferson St. bridge, the additional strength provided by the CFRP system was:

M,;trp = 1,027 k-ft > Mreqa’ = 1,000 k-ft (ok)

For the girders in the 19th Avenue structure the additional strength provided by the CFRP was: M,;trp = 392 k-ft > Mreqa’ = 360 k-ft (ok)

Additionally, ADOT requested that a protection system be implemented to prevent concrete fragmentation due to spalling and/or impact of the girders. Bi-directional CFRP installed in a U-wrap scheme was initially decided and detailed as the best solution to use as a “net”, having strength in both directions. The 6-inch long carbon fiber tows in the bi-directional fabric are too short to properly develop strength and be effective. Furthermore, the short fibers make the installation of these strips and maintaining the alignment of the fibers very difficult. As the discussions with ADOT progressed, ADOT determined that it was in their best interested to use 6-inch wide unidirectional carbon strips (QuakeWrap® VU27C) with a 12 inch center to center spacing of the strips. After the first day of construction, ADOT agreed with the suggestion to switch the U-wrapped fabric to a uni-directional fabric and that is what was installed on both bridges.

Typical details of the CFRP installation drawings are provided in Fig. 5. All the CFRP fabric used was QuakeWrap® TU27C. The extent of the retrofit was limited to the regions indicated by ADOT; these regions were between 40′ – 50′ in length.

CONSTRUCTION PHASE/FIELD INSTALLATION

The construction phase of the CFRP repairs on ADOT project PHX-Cordes Jct. Hwy (1-17: 19th Ave. and Jefferson Street Bridges) was put out to bid and then awarded to FNF Construction out of Tempe, AZ. FNF Construction is a heavy civil general contractor with extensive experience working with ADOT, but lacking experience in the installation of CFRP systems. They subcontracted with FRP Construction LLC. to perform the strengthening system installation . This was to be the first CFRP repair done for ADOT and it required in depth work in all aspects of the project. From the planning through the execution phase, this project was unique in a variety of aspects.

During the planning phase, the project manager for FRP Construction was required to attend a weekly status meeting with FNF Construction and ADOT. These meetings included the review of material submittals, design modifications, and construction sequence scheduling. Before the project was allowed to be executed, every item was discussed in detail. All information had to be relayed to ADOT and accepted before any construction work could begin. These status meetings took place 1 month before the project was to be executed. Being that this was ADOT’s first CFRP repair, there were many questions to be asked and specifications to be modified. FRP Construction and its engineering partners QuakeWrap I nc. were able to successfully assist ADOT in providing adequate specifications that were feasible for a company to adhere to.

On April 24, 2017, FRP Construction began the CFRP install on the bridge girders. This work could only be done at night due to traffic restrictions. The available hours for work completion were from 9:00 P.M. to 5:00 A.M. The travel lanes of Interstate 17 were closed each night and reopened at exactly 5:00 A.M, to ensure traffic was not restricted during high vehicle volume use. A 6-man crew from FRP Constru ction performed this work within these time constraints. In addition to FRP Construction’s crew on site, FNF Construction provided a crew for traffic mitigation, concrete repairs and a final painting finish. Also, ADOT’s quality control team and design engineers made frequent visits to the sites for review and documentation purposes.

The location of the single damaged girder on Jefferson St. bridge and the installed flexural CFRP reinforcement is shown in Fig. 6.

The full scope of the project also included repairs to an additional 11 girders on the 19th Ave. bridge (fig.7). The design documents provided by QuakeWrap Inc. detailed several types of repairs and which girders receive them. Some girders received both longitudinal fabric with U-wrap strips, while others just received the U-wrap strips.

The first day of CFRP construction efforts began at a slow pace. At the end of the first shift the expected progress was not met. FRP Construction knew one of the steps in the process was consuming a larger amount of man power than expected. The U-wrap strips were required on every girder in the retrofit plan and originally detailed by QuakeWrap using bi­ directional CFRP fabric. Cutting the TB20C biaxial fabric into small strips for the install proved to be a task that would likely exceed the time allotted to complete the project. FRP Construction requested that this fabric be cut off site in between shifts but was denied by ADOT due to quality control measures being used on site. In light of the time constraints, FRP Construction sent an RFI to QuakeWrap Inc. and obtained a design change that was approved by ADOT to install TU27C uni­ directional fabric as the 6-inch U-wrap strips. This fabric was much easier to cut and therefore quicker to install for the crew. With only one man-lift at our disposal, FRP was able to cut and install the fabric at a rate of about 500 SF per night.

FNF Construction scheduled the project to take approximately 20 working days for the fabric installs on the 19th Ave. bridge and requested that the Jefferson bridge be retrofitted over the span of one weekend starting Friday night at 9pm and ending Monday morning at Sam. FNF and ADOT were both very pleased to find the 19th Ave. bridge completed in approximately 15 days and the Jefferson St. bridge completed over the single weekend in May of 2017.

Quality controls for this project were specified by ADOT and added upon by FRP Construction. The materials provided by QuakeWrap Inc. were required to arrive on site in unopened packaging. The materials are identified and recorded by Lot numbers for the fabric and Batch numbers for the epoxy resin. FRP Construction, of their own accord, provided ADOT with a lay out of Lot and Batch numbers for each girder.

The project specifications required other quality controls using typical testing methods. FRP Construction subcontracted with Western Technologies of Phoenix, AZ to provide third party testing results to ADOT for the entire project. Multiple testing samples called witness panels, were made every day. One witness panel per day was required to be tested by Western Tech. The testing was performed using the ASTM standard D3039 and all the results were sent to QuakeWrap for verification. QuakeWrap assembled the results and submitted them to ADOT. Additionally, a couple of representative areas were prepared and tested for bond strength to the concrete substrate using ASTM C882 to guide the efforts. Those results were requested by ADOT’s QA representative even though they were not required per the published specifications and they were performed by FRP Construction.

FRP Construction planned and executed this project with the same commitment to quality that it brings to every project. Their attention to detail and safety have contributed to the unprecedented growth rate of this company over the years. The expertise and experience each employee brings to the table have helped FNF Construction deliver to ADOT another successfully completed project. Future endeavors of ADOT using CFRP have been strengthened, thanks to the efforts and experiences shared during this first project.

SUMMARY AND CONCLUSIONS

A total of 12 AASHTO Type II girders were repaired using CFRP products. One of the girders was severely damaged by a truck collision, resulting in a 50% loss of strength. The other 11 girders were not as badly damaged, resulting in a loss of 20% of flexural capacity. The repairs were successfully completed on schedule during evening hours and with minimal disruption to traffic . This positive first application of CFRP by ADOT will inevitably result in a wider use of this technology in Arizona and elsewhere.

ACKNOWLEDGEMENTS

The repairs discussed above were carried out by FRP Constru ction, LLC, Tucson, Arizona as a subcontractor to FNF Construction, Tempe, AZ.

REFERENCES

American Association of State Highway and Transportation Officials – Guide Specifications for Design of Bonded FRP Systems for Repair and Strengthening of Concrete Bridge Elements (1st Edition), AASHTO FRPS, 2012.

American Concrete Institute – Committee 440. Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures, ACI 440.2R-08, ACI, Farmington Hills, MI, 2008.

Ehsani, M.R. and Saadatmanesh , H. “Fiber Composite Plates for Strengthening Bridge Beams,” Composite Structures, 1990, 15(4), 343-355.

DOWNLOAD PDF

IBC-16-19

KEYWORDS: Bridge pile encasement, confinement, fiber reinforced polymer (FRP), underwater pile jacket, alkalisilica reaction (ASR), asset management, prestressed concrete, rehabilitation, maintenance, Barron River bridge

ABSTRACT

Octagonal PSC piles in a bridge over the crocodile-infested Barron River near Cairns in tropical north Queensland, Australia, experienced severe ASR cracking and required remediation to protect against corrosion damage. Forty piles were repaired underwater by encasing in a protective jacket consisting of thin glass FRP laminate sheets wrapped around the piles to create a seamless, impervious, cylindrical shell subsequently filled with a low viscosity resin that sealed the concrete and filled any voids and cracks.

INTRODUCTION

The bridge on the southbound carriageway of the Captain Cook Highway over the crocodile-infested, tidal estuary of the Barron River 4.5 miles (7km) north of Cairns in tropical north Queensland, Australia, was constructed in 1977.

The bridge is 579 feet (176.5m) long and 34 feet (10.4m) wide, with seven spans of 77 feet (23.5m). The superstructure consists of prestressed concrete (PSC) beams and reinforced concrete (RC) deck, while the substructure consists of piled abutments and blade piers rising from pile caps, each supported on 10 closely-spaced PSC piles.

Piles are 20.7 inches (525mm) octagonal PSC piles. All piles are raked with clear spacing between piles as low as 12 inches (300mm).

The top of the pile cap is nominally at mid-tide. Tidal range is around 15 feet (4.5m) and the upper 4 feet (1.25m) of piles is within the tidal range. However, the piles are generally underwater and during normal tidal movements, it is rare for more than 20 inches (0.5m) to be exposed (Fig. 1).

In 2000, severe ASR cracking was identified in 13 of the 40 PSC piles in the four piers located in the main river channel.

To protect the piles from premature failure due to corrosion, and to inhibit the development of ASR cracking, the state road authority – the Queensland Department of Transport and Main Roads (DTMR) – determined that encasing the piles was the preferred long-term treatment.

This paper describes a number of phases associated with a 15-year journey in search of a suitable, economical pile encasement solution, and describes details of the adopted solution and its installation.

EARLY CONSIDERATIONS

Initial preference was to use a traditional RC jacket to protect the piles. However, further investigation identified a number of site challenges to traditional encasement solutions, including:

The close-spacing of piles precluded the use of individual RC encasements and increased the complexity and cost of an RC encasement considerably.

Having utilized Fiber Reinforced Polymer (FRP) jackets in above-water pile remediation works, DTMR was interested in exploring the feasibility of utilizing FRP jackets in underwater environments. Sample of FRP jackets that were common for pile encasement projects are shown in Fig. 2. There are several inherent problems with these products:

• The vertical seam along the side of the jacket allows ingress of moisture and oxygen so the corrosion process will continue;
• The vertical seam and bolts introduce a plane of weakness in the jacket that limit the confining pressure exerted by the jacket to the swelling pile; and
• The jackets must be manufactured for a specific size; this delays project scheduling and construction time

FRP PROTYPE DEVELOPMENT (2003)

A review of available FRP jackets failed to identify a commercially-available jacket that met the desired technical requirements and could be easily and confidently installed underwater.

DTMR became an active partner with the University of Southern Queensland (USQ) and the NSW Roads and Traffic Authority (RTA) in a joint project to develop a prototype FRP jacket for concrete piles suffering ASR cracking.

Various researchers, including Carse (1992), had shown that the rate of ASR reduced as internal pressure increased. One of the primary goals, therefore, was to develop an FRP jacket that would not simply act as a barrier to inhibit the ingress of chlorides, moisture and oxygen, but would provide sufficient confinement to allow a buildup in internal pressure to slow the ASR reaction. A practical operating pressure of 2MPa was adopted for purposes of design of a confinement system (Fig. 3). As discussed above, the FRP jackets that were in use at the time could not provide an impervious moisture barrier nor could they resist a confining pressure of 2 MPa (300 psi).

A prototype FRP jacket was subsequently designed and tested. At the time, it was considered that underwater-curing epoxies were not sufficiently advanced to meet the project objectives, and a mechanical jointing system was preferred. The prototype was designed with a finger joint system, with the fins being turned in to be anchored in the hardened grout subsequently placed in the annulus between the pile and jacket (Fig. 4).

Testing of the prototype was partially successful, proving the concept, but requiring further development to progress to a commercial product (McGuffin et al. 2003).

Unfortunately, the project did not progress beyond the prototype phase.

MARKET EXPLORATION (2006)

In late 2006, there was a call for Expressions of Interest (EOI) for the jacketing of the piles. Encouragement was given to offer innovative jacketing solutions including FRP jackets.

From four submissions, all proposing RC encasement, two companies were selected to develop and provide an indicative price for their proposed solutions.

Though one option initially appeared to offer significant cost savings, further development taking into account the specific challenges of the site, resulted in a final estimate well in excess of the available funding.

At this stage, it was decided to:

PROJECT REINITIATION (2014)

In 2014, all piles were cleaned and inspected by divers at a total cost of A$150,000 – 50% of which was for cleaning, 30% for installation of crocodile protection and 20% for inspection.

The underwater inspection indicated that the number of cracked piles had increased from 14 in 2004 to 22 in 2014.

At the same time, the recently developed PileMedic® system was identified as a pile remediation system that might be capable of meeting the technical and practical needs of this project at an economical cost (around 40% of the previously tendered prices for a traditional RC encasement).

A project proposal was prepared and subsequently ratified as a trial installation by DTMR’s Structures Branch. Finally, funding was secured, design details finalized and scheme documents prepared to enable the works to proceed.

DTMR’s in-house delivery arm (Roadtek) were engaged to undertake the works, sub-contracting QuakeWrap (Australia) to provide training and technical advice during the works and local commercial divers to undertake underwater works.

THE ADOPTED SOLUTION – PILEMEDIC®

Independent from these developments in Australia, a new form of FRP product was being created by one of the co-authors in the U.S. (Ehsani 2009). This development had its beginning dating some 25 years earlier when FRP products were being studied for strengthening of bridge piers to resist seismic loads (Saadatmanesh et al. 1992). The early generation of FRP was applied using a technique known as wet layup. In that system, fabrics of carbon or glass are saturated with epoxy and directly wrapped (i.e. bonded) to the column for increased confinement. While that system does provide structural enhancement, it has a few limitations. The surface of the column must be repaired and patched to bring it to a smooth surface prior to wrapping of fabric. Furthermore, wet layup systems require immediate bonding to the surface of the column and are difficult to implement when a column is submerged in water — as was the case here.

The new PileMedic® laminates are constructed with specially-designed equipment. Sheets of carbon or glass fabric up to 60 inches (1.5 m) wide are saturated with resin and passed through a press that applies uniform heat and pressure to produce the laminate (Figs. 5 and 6). The laminates offer three major advantages compared to other FRP laminates. First, by using a combination of unidirectional and/or biaxial fabrics, the laminate may provide strength in both longitudinal and transverse directions; this is a tremendous advantage that opens the door to strengthening of columns for both bending and confinement. Secondly, they are very thin; with a typical thickness of 0.025 inches (0.66 mm), they can be easily bent into any shape circle with a diameter of 12 inches (300 mm) or smaller (Fig. 6). Lastly, the number and pattern of the layers of fabrics can be adjusted to produce an endless array of customized products that can significantly save construction time and money.

The fiberglass jackets depicted in Fig. 2 are basically a stay-in-place form for placement of grout. They offer virtually no structural benefit to the pile. For example, the tensile strength of some of the most popular fiberglass jackets is compared to PileMedic® laminates in Fig. 7.

The tensile strength of PileMedic® laminates ranges between 62 to 156 ksi (427-1080 MPa). The thickness of the laminates ranges from 0.01 to 0.03 inches (0.25 to 0.76 mm). It is noted that the development of an equipment to produce such uniformly thin and wide laminates is not a trivial matter and has been a major part of this product development.

The first application of PileMedic® for repair of submerged piles was carried out in Miami, Florida(Ehsani and Tipnis 2011). Details of the installation can be viewed in this video. Since that time, many transportation agencies have performed extensive testing of this product for various bridge pier applications. Among these are studies funded by Caltrans for concrete columns damaged in earthquakes (Yang et al 2015), Texas DOT for corroded steel H piling (Dawood et al 2015), and Nebraska Dept. of Roads for deteriorated timber piles (Mohammadi et al 2014). All of these studies have demonstrated the structural enhancement that can be obtained by using these laminates.

DESIGN OF THE PILE JACKET

When the objective of the jacket is to provide confinement for the column or pile, the design is explained with the aid of the free body diagram shown in Fig. 8. Assuming an element of the column with unit height, let us assume that the diameter of the jacket that encompasses the pile and the grout in the annular space is 25 inches (635 mm). If the jacket is made by wrapping two layers of PileMedic® PLC150.10 laminates, each one inch width of this laminate has a breaking strength of T= 4050 pounds (18 kN). The confining pressure, p, acting on the pile is the reaction of the sum of the 4 tension forces shown in Fig. 8, such that: 4T = p x Jacket Diameter, or p= 4T/Diameter = 4 x 4050/25= 648 psi (4.38 MPa).

As demonstrated, the laminates provide significant confining pressure for the pile. This confining pressure increases the compressive strength of the original pile and the newly placed grout. The amount of this confining pressure can be changed by using 0 40,000 80,000 120,000 160,000 Tensile Strength (psi) different types of laminates that have different tensile strengths, and by changing the number of layers that are wrapped around the pile. It is noted that this example is only intended to demonstrate the concept of the design. Safety factors and durability considerations must also be included. Design Guidelines such as those published by ACI Committee 440 provide detailed information for such factors.

In some applications, the design may call for an increase in the bending capacity of the column as well. Laminates constructed with biaxial fabrics that provide strength in hoop and longitudinal direction (along the height of the pile) are ideal for such cases. By using an epoxy grout, the jacket will be bonded to the concrete pile and it will contribute to the bending capacity of the pile or column. The Caltrans test mentioned earlier was one such application.

FIELD INSTALLATION

A total of four piers each having 10 submerged piles were repaired on this project. The bridge site on Barron River is known for its crocodiles, bull sharks, jellyfish and the like. For safety of the diving crew, a steel cage was positioned around each group of pier (Fig. 9).

A barge was used to provide a flat working platform and a staging area for the materials and equipment (Fig. 10). The barge also included a small crane for loading of the materials such as the resin and grout.

PileMedic® laminates are supplied in rolls that are 4 feet (1.2 m) wide and 250 feet (76 m) long. A major advantage of this product is that the rolls can be easily cut to any length in the field, allowing a single roll of laminate to be used for repair of piles of virtually all shapes and sizes. Standard details call for the laminate to be wrapped twice (i.e. 720 degrees) around the pile with an additional 8 inch (200mm) extension beyond the starting point.

The repairs required creating a 2 inch (50mm) annular space around the piles to be filled with grout. This meant a diameter of 25 inches (635mm) for the jackets. Such jackets would be small enough to leave adequate working space between adjacent piles during the repair.

The circumference of such a jacket is 78.5 inches (2m). So, the 4-ft (1.2m) wide laminates were cut in lengths of 2×78.5+8 = 165 inches (4200mm) in the field. The laminate is placed flat on a table (Fig. 11). A two-part epoxy resin paste is supplied as a part of this system. This resin cures underwater and Figure 10 – Barge used as working platform eliminates the need for creating cofferdams and dewatering around the piles.

The laminate will be wrapped around the pile to create a two-ply shell. In doing so, the portion of the laminate that is adjacent to the shell does not have to be coated with epoxy. But the second ply and the 8 inch extension, must be coated. In this case, a length of 78.5+8=86.5 inches (2200mm) of the laminate had to be coated with epoxy. The twocomponent epoxy has a consistency similar to tooth paste and can be applied using a notched trowel to achieve a uniform thickness of 30-40 mil (0.75-1 mm).

Next, the coated laminate is picked up and passed over the steel cage to the diving crew (Fig. 12). In spite of their size, the lightweight jackets are fairly easy to handle once they are submerged. Spacers such as short pieces of 2 inch (50mm) diameter PVC pipe can be attached to the pile surface to define the required annular space for grout. The crew wraps the laminate around the pile over the spacers. The epoxy paste serves as a lubricant and allows easy adjustment of the size of the jacket. Ratchet straps can be wrapped around the shell to maintain its shape and dimension while the epoxy cures.

The first 4-ft (1.2m) tall jacket was installed at the base of each pile and extended 18 inches (450 mm) below the mud line. The next jacket was applied with a 4-inch (100 mm) overlap along the height of the pile. An epoxy paste is applied to this overlapping region. The process continued until a jacket of desired height was created. The diving crew consisted of 4 workers: 2 divers performing the installation, 1 back up diver and 1 supervisor. On the barge, a two-man crew was given the task of cutting the laminates and mixing and applying the epoxy paste.

Once the jacket was installed, the annular space was filled with a cementitious underwater grout using the tremie placement method. A shear mixer and pump were utilized for this application. As the grout is placed, its hydrostatic pressure pushes the two layers of the laminate against each other for a perfect bond. Moreover, the heat of hydration of the grout can also assist with the faster curing of the epoxy paste. In the following day the ratchet straps will be removed and the repair is complete. Figure 13 shows the completed project after all 40 piles were successfully repaired.

The field installation had its unique challenges. The repairs were scheduled for the dry season to avoid flash floods in the river. The visibility in the murky water was low and often limited to 1 ft (300 mm). The large tidal flow ranging 6-10 feet (2-3 m) also had to be considered.

The project was completed in around 5 working weeks (35 days). This time was divide approximately into one day to install the steel protective cage, four days to wrap the piles and place the grout and one day to demobilize from one pier and move to the other. The crew performed the work exceptionally well, adapting to the constant challenges faced working in a trying tropical environment.

COST SAVINGS

According to the data from DTMR, the following cost data in Australian Dollars were reported. The original concrete encasement option in 2008 was bid at $2.0M which is equivalent to $2.46M in 2015. The PileMedic® option was estimated at $1.1M, but the actual cost ended up being only $0.95M. The $150,000 reduction was primarily due to ease of installation of the system and the efficiency of the diving crew as the project progressed.

The adopted solution resulted in savings of $1.5M or 160% compared to the original reinforced concrete encasement option. In retrospect, the agency’s delay of the project turned out to be a good decision as the development of this new technology resulted in significant cost savings.

SUMMARY AND CONCLUSIONS

The repair of 40 ASR contaminated submerged piles had lingered on for nearly 15 years. The options for a suitable repair to provide an impervious structural shell that would both stop the ingress of water and provide high confinement for the piles to resist the swelling of the concrete were limited at best. During this period a new product was developed and introduced to the market that could achieve both of these objectives.

The repairs were completed in 2015 with 160% cost savings compared to the original solution that was considered. This brought a happy ending to the 15 year search to solve this challenging problem.

ACKNOWLEDGEMENTS

This project was a truly collaborative effort that involved many agencies in two continents. The authors would like to thank the contributions of the Structures Branch and far North District of TMR, Roadtek as the General Contractor and JD Marine for underwater works.goo

The PileMedic® system and the method of repair presented are protected by U.S. Patent No. 8,650,831 and other pending U.S. and international patent applications.

REFERENCES

American Concrete Institute – Committee 440. Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures, ACI 440.2R-08, ACI, Farmington Hills, MI, 2008.

Carse, A. Report on the mechanism causing cracking in the prestressed piles of the Houghton highway bridge, Queensland Department of Transport Technology Division, 1992.

Dawood, M., Karagah, H., and Shi, C. “Repair Systems for Deteriorated Bridge Piles Technical memorandum #4a,” TxDOT Project 0-6731, University of Houston, TX, 41 pp, May 2015.

Ehsani, M. and Tipnis, A. “Miami Building First to benefit from Innovative Pile Jacket,” Concrete Repair Bulletin, 16-18, July-August 2011. (Video)

Ehsani, M. “FRP Super Laminates Present Unparalleled Solutions to Old Problems,” Reinforced Plastics, 40-45, 53(6), 2009.

McGuffin, J., Heldt, T., Van Erp, G and Cattell, C. “Feasibility of a new pile rehabilitation system” Fibre Composites Design and Development, USQ Publication No FCDD-0300019_R3, 2003.

Mohammadi, A., Gull, J., Taghinezhad, R., and Azizinamini, A., “Assessment and Evaluation of Timber Piles Used in Nebraska for Retrofit and Rating,” Draft Final Report to Nebraska Department of Roads, Dept. of Civil Eng., Florida International U., Miami, FL, 51 pp, April 2014.

Saadatmanesh, H., Ehsani, M.R., and Li, M.W. “Seismic Strengthening of Concrete Columns with Fiber Composite Belts,” Proc., ASCE Materials Eng. Congress, Atlanta, August 1992, 677-690, 1992.

Yang, Y., L. Sneed, M. Saiidi, A. Belarbi, M. Ehsani, and R. He. 2015. “Emergency Repair of a RC Bridge Column with Fractured Bars using Externally Bonded Prefabricated thin CFRP Laminates and CFRP Strips,” Composite Structures, Elsevier, 727- 738, 133, 2015. (Video)

DOWNLOAD PDF

Abstract: Hollow steel tubes (HSS) are frequently used in electrical substations to support various equipment. Even though they are made of galvanized steel, poor details and/or deficiencies during the galvanizing process can result in corrosion due to the accumulation of moisture inside the base of these tubes. Unfortunately, the deterioration is usually concentrated at the base of the structure columns, where the stresses are at maximum. If repairs to the columns are not performed, the structural integrity is compromised.

This paper reports on the newly developed Fiber Reinforced Polymer (FRP) SuperLaminate™ and its use in repair of nearly sixty corrosion-damaged structures at two Tucson Electric Power (TEP) substations. The laminates are supplied in 4-ft wide rolls that are approximately 0.025 inch thick. The tensile strength of these laminates varies from 60,000 to 150,000 psi depending on whether they are constructed with glass or carbon fiber. In the field, the laminate is cut to the desired length and coated with an epoxy paste. The laminate is wrapped around the column a minimum of two wraps, creating a 4-ft tall cylindrical shell around the column. The annular space between the column and the shell is filled with concrete or grout. Additional reinforcing steel can be added in the annular space. The lower portion of the column is also filled with grout and a drain is added to prevent future accumulation of moisture. All repairs can be performed while the structure remains in service.

Introduction

During an inspection at one of Tucson Electric Power’s substations, extensive corrosion was discovered on many of the substation support columns. Based on a visual inspection, it seemed that the corrosion had begun on the inside of the column wall meaning that the damage was potentially much worse than it appeared. An ultrasonic thickness gauge was used to determine the steel thickness adjacent to the visible damage. Measurements along the height of the columns indicated that the loss of wall thickness was more severe closer to the base of the column (Fig. 1). These investigations also revealed that in many cases the majority of the structural damage was on the inside of the square columns with little or no visible damage from the outside.

Since the majority of the structural damage was found to be on the inside of the square columns, it was concluded that the damage is most likely the result of deficiencies during the galvanizing process. It is likely that the acid bath was not properly drained from the columns during the prepping stage causing the inside of the square columns near the base not to be galvanized properly. In addition, during the installation of these columns, grout had been placed between the base plate and the foundation. This caused rainwater to be trapped inside the structure, which accelerated the corrosion process.

Earlier attempts to repair these columns included the use of steel collars as shown in Fig. 3. These collars had to be manufactured to fit the various size-damaged columns. The steel collars were bolted to the columns. An examination of these repaired columns showed that the corrosion process still continued inside the columns and the steel collars would hide the damage. In essence this would offer a false sense of security. These collars were also very heavy and difficult to install. They had to be custom designed to ensure that they could be installed considering the base configuration at each column and the supporting equipment that may be present.

These finding resulted in Tucson Electric Power seeking a better and more permanent solution for repair of these columns. Among the repair techniques, was a recently developed jacketing system that used Fiber Reinforced Polymer (FRP) products.

FRP Products for Repair of Structures

FRP products had been introduced in the late 1980s for repair and strengthening of existing structures (Ehsani & Saadatmanesh, 1990). The original technique for application of these is know as wet layup. In the wet layup process, rolls of fabric made with carbon or glass fabric are saturated with epoxy resin and are externally bonded to the surface of the structure requiring repair or strengthening. The materials cure in a few hours and become very strong in tension. Typical tensile strength for FRP products are 2-4 times that of steel. In addition to high tensile strength, the durability of these products that do not corrode and the ease of installation has made them very popular for repair and strengthening of civil infrastructure. Numerous buildings, bridges, pipelines, etc. have been retrofitted with these products worldwide. With the publication of design guidelines (ACI, 2008) it is fair to say that FRP is no longer an experimental product and it is rapidly gaining acceptance in the repair and retrofit industry.

The original form of FRP products that have been used to date are primarily. Fabrics offer the widest versatility in the field and are installed following a procedure commonly referred to as the wet layup method. This technique requires properly trained technicians to prepare the resin in the field, saturate the fabric with the resin and apply it to the structural member. Care must be taken to ensure that the fibers of the fabric are aligned in the correct direction and to remove all air bubbles before the fabric is cured. As a result, the quality of the finished FRP product is greatly influenced by the experience of the installation team. However, for these corrosion-damaged steel columns, FRP fabrics do not offer a suitable solution.

The author has recently developed a new type of FRP products called SuperLaminate™ that could be of great value for certain repairs such as these steel columns (Ehsani 2010).

Super laminates are constructed with specially-designed equipment. Sheets of carbon or glass fabric up to 60 inches (1.5 m) wide are saturated with resin and passed through a press that applies uniform heat and pressure to produce the laminate (Figs. 4 and 5). Super laminates offer three major advantages over conventional laminates. First, by using a combination of unidirectional and/or biaxial fabrics, the laminate may provide strength in both longitudinal and transverse directions; this is a tremendous advantage that opens the door to many new applications (Table 1). Secondly, they are much thinner than conventional laminate strips; with a typical thickness of 0.025 inches (0.66 mm), they can are flexible enough to be formed into various shapes in the field. Lastly, the

Type of Fiber	Thickness in. (mm)	Tensile Strength KSI (MPa)	Tensile Modulus KSI (MPa)
Carbon	0.026 (0.66)	156 (1,080)	13,800 (95,500)
Carbon	0.026 (0.66)	101 (698)	7,150 (49,280)
Glass	0.026 (0.66)	62 (431)	3,500 (24,140)
Glass	0.011 (0.28)	49 (335)	3,200 (22,060)

number and pattern of the layers of fabrics can be adjusted to produce an endless array of customized products that can significantly save construction time and money.

Field Application

After investigating various alternatives, TEP chose the newly-developed super laminate FRP system for repair of the structures that could be implemented while the structure was in service and resulted in continuing system reliability. With proper oversight of safety personnel, the structures could be repaired without need for a costly outage.

The repair of the columns with laminates was aimed at strengthening the lower 3 feet of the structures. First, the corroded material that had separated from the inside of the structure was removed and the column was cleaned of any rust (Fig. 6a).

An access port about 3 inches in diameter was cut at an elevation of 33 inches from the base. The openings were temporarily sealed with plywood and clamps. A non-shrink high-strength grout was mixed and pumped into the structure from the access port (Fig. 6b). Four No. 5 U-shaped reinforcing bars were welded to the base plate; these bars were primarily for enhanced flexural capacity of the structure.

Due to its dielectric properties, a glass laminate with tensile strength of 62000 psi and thickness of 0.026 inch was used. A 3-foot wide x 8.5-foot long piece of laminate was used for each structure (Fig. 6c). A special two-component epoxy was mixed in the field and applied to an approximately 5-foot long section of the laminate. The mixed epoxy has a paste-like consistency and is applied with a trowel to a thickness of 20-30 mil (Fig. 6c). The laminate is then wrapped loosely around the structure to create a two-ply shell (Fig. 6d). The 8.5-foot length of the laminate allows creation of a two-ply 15-inch diameter shell with 8 inches of overlap beyond the starting point. At this stage, before the epoxy cures, ratchet straps must be used to hold the shell in the desired shape.

A 1-inch diameter PVC drain pipe was installed to make sure no rainwater will accumulate inside the structure (Fig. 6e). The bottom edge of the jacket was sealed with tape atop the base plate.

The annular space between the jacket and the steel column was also filled with the same non-shrink grout (Fig. 6f). Consolidation of the grout was completed with a small vibrator and the top of the grout was finished with a trowel. Before the grout sets, the hydrostatic pressure from the grout pushes the inner layer of the jacket outward against the outer layer and forces the two plies of the laminate to be tightly pressed against each other.

After several hours, the ratchet straps were removed and the exterior of the shell was painted with a UV-protecting coating (Fig. 7). All of these repairs were performed while the substation remained fully operational.

In this particular application, the intent was to restore the original capacity of the structure and therefore the foundation did not require any strengthening. In cases where a strengthening of the steel structure beyond its original capacity is desired, the capacity of the base plate and foundation must also be checked. It may be necessary to use anchor bolts to improve the overturning capacity of the foundation. Such an approach has been used recently to repair monopole towers that are used in cellular phone communication industry (PileMedic 2104).

Summary and Conclusion

The original repair of these structures that used steel collars allowed presence of moisture inside the columns which would lead to continued corrosion and degradation of the structure. The repaired structures using the FRP system now exceed the design strength capacity. This repair system eliminated the possibility of water accumulation inside the columns and the corrosion problem has been completely eliminated on the repaired Fig. 7 Completed repaired column structures. A further primary advantage of this repair system is that no electrical outage was required during the repair process.

Acknowledgements

The repair technique described in this article is protected by US Patent 8,650,831 and other pending U.S. and international patent applications.

References

ACI Committee 440. 2008. Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures, American Concrete Institute, 440.2R-08.

Ehsani, M.R. and Saadatmanesh, H. (1990). “Fiber Composite Plates for Strengthening Bridge Beams,” Composite Structures, 15(4), 343-355.

Ehsani, M. 2010. FRP Super Laminates: Transforming the Future of Repair and Retrofit with FRP,” Concrete International, ACI, 32(03): 49-53.

PileMedic. 2014. “Strengthening of Cell Phone Communication Towers with PileMedic®” , A video available on www.YouTube.com.

DOWNLOAD PDF

Most structural engineers frequently encounter the design of buildings and bridges in their careers. However, a relatively small fraction get involved with the design of marine infrastructure such as submerged piles, see walls or sheet piles. This paper summarizes the structural enhancements and improved constructability offered by a new system developed by the author; full details of the system are available in two recent webinars that have been recorded and area available for download at www.PileMedic.com/webinars.html

PileMedic® is based on a new type of Fiber Reinforced Polymer (FRP) SupperLaminate developed a decade ago (Ehsani 2010). These laminates are made in a special process where one or more layers of reinforcing fabric are saturated with resin and subjected to heat and pressure to produce a very thin laminate. The laminates are typically supplied in rolls that are 4 feet wide by hundreds of feet long. The main advantages of the laminate are described below:

1. Size & Shape – The small thickness of the laminates which is typically less than 0.05 inches makes them flexible enough to be wrapped around piles of any shape and size. Standard detail requires wrapping the laminate twice (i.e. 720 degrees), plus an 8-inch extension around the pile. Once the right length of laminate is cut from the roll in the field, the second half of the laminate is coated with an epoxy paste, and it is wrapped around the pile and bonded to the first layer, creating a two-ply structural shell around the pile (Fig. 1). This eliminates the need for custom ordering the jackets, saving significant time, and shipping and storage cost. The annular space between the jacket and the pile will be filled with grout, concrete or epoxy grout later.
   
   
   Fig. 1. Laminate being cut to size and coated with epoxy on a barge; divers wrapping the laminate around a pile after the reinforcing bars have been secured in the spacers.
2. Confinement – Depending on the type of fiber that is used, carbon or glass, the laminates have tensile strength ranging between 28,000 and 150,000 psi. Furthermore, by selecting the orientation of the fibers within the fabrics the strength of the laminate in the longitudinal and transverse directions can be adjusted to meet the specific project requirements. The jacket creates a uniform confining pressure (i.e. 360 degrees) around the pile and it eliminates the need for hoop reinforcement. This allows divers to install the vertical reinforcing bars individually (when needed). The elimination of handling of assembled reinforcing cages that requires two divers, results in significant savings for the project.
   
   
   Fig. 2 Jackets provide uniform confinement around the pile
3. Corrosion Protection – It is well recognized that the oxygen present in water is the fuel for the corrosion process. The laminates create a seamless shell around the pile, preventing any moisture or oxygen ingress. This will drastically reduce the corrosion rate in the pile.
   

As part of the development of this repair system for use by the U.S. military and other clients, several accessories have been produced that significantly enhance the quality of construction and result in more accurate engineering design. These are described below:

• Spacers – These plastic parts are available in a variety of shapes and sizes and serve multiple purposes (Fig. 3). Divers can pass a zip tie through several spacers and fasten them around the pile. The longitudinal reinforcing bars (typically made of glass FRP) can be snapped into place (Fig. 1). This allows the engineer to know the precise location of each bar. The overall length of the spacers defines the annular space between the jacket and the pile that will be filled with grout or concrete. A variation of these spacers can be custom made to attach to the flanges of H-piles for placement of the longitudinal bars. In some cases, the repair portion of the pile extends into the mud line and the soil provides a seal at the bottom of the jacket for placement of the grout. However, there are many applications where the length of the repair is limited to the splash zone, requiring the creation of a seal at a point along the height of the pile where the bottom of the jacket will be positioned. The skirt pin shown in the foreground of Fig. 3 allows the creation of that seal.
  
  
  Fig. 3. Sample spacers and skirt pin.
• Shear Transfer – In repaired piles, a load path must exist for the forces in the newlyformed reinforced concrete shell to bypass the damaged zone of the pile. When the pile is made of timber or concrete, the rough surface of the host pile usually provides sufficient load transfer through bond. However, when the host pile is made from steel, a mechanism such as welded shear studs can be used. Underwater welding is costly and difficult and as was the case on one of our recent projects, the flammable materials present on the site did not allow any welding. Two newly developed products assist with this shear transfer. One is called ShearWrap™ for use on circular steel piles. Each band is about 3-4 inches wide, 0.2 inches thick and is comprised of two half circles. The divers place the band around the pile (usually above water) and loosely connect the bolts. The band is then free to slide along the height of the pile to its final position. At that point, the divers tighten the two bolts to the specified torque. In some cases, wings can be welded to the steel bands to increase the bearing capacity of the unit. Similarly, for steel H-piles, ShearClamp™ can be used. These individual clamps include a bolt that can be tightened against the flange of the pile. The capacity of each unit can be determined based on its resistance through friction and bearing.
  
  Fig. 4. ShearClamp™ used on H-piles and ShearWrap™ steel bands that are torqued around circular steel piles.

Tests:

The PileMedic® system described above has been tested by several government agencies. Caltrans conducted tests of concrete bridge piling which were severely damaged in an earthquake. It was shown that even when the longitudinal reinforcing steel in these columns were fractured, the full capacity of the column could be restored in 48 hours without the need for replacing the broken reinforcing steel. Texas DOT conducted a study of corroded steel H-piles at the University of Houston. Even sections with up to 80% section loss gained their full strength after being repaired with this system. Nebraska Department of Roads studied repair of timber piles.

More recently, the US Army Corps of Engineers (ACE) conducted a major investigation of solutions for exigent repair of piles. Based on full-scale tests, ACE has selected PileMedic® as the only solution for the US military worldwide to repair timber, concrete and steel piles. As a part of this evaluation, 90 concrete piles were repaired at Pearl Harbor Hawaii. The installations also considered various conditions such as poor visibility in the water and water current. As noted by some of the diving team members, the new system reduces the repair time to as little as 1/3 of the previously used pile jackets. Considering that the labor/diving fees account for some 75% of the total repair costs, the new system could result in significant time and cost savings in pile repair projects.

Field Installation

The system described here has been installed in many projects worldwide. Among these are over 500 timber piles at Perdue Agrobusiness in Chesapeake, VA, 90 concrete piles for the Port of Seattle, 95 steel piles in Houston, 265 steel piles for Indiana DoT, 100 piles in a jetty belonging to the Nigerian national Petroleum Corporation, numerous bridges in Australia, etc. A related application of this system is for repair of cell phone towers. As the communication companies need to add more equipment to existing towers, these towers that are often placed in congested areas must be strengthened. Some 100 such towers have been retrofitted in California alone with this proprietary technique.

Videos showing the highlights of many of these projects are available on our YouTube channel and can be viewed at this links: https://tinyurl.com/y8mvdhs9

DOWNLOAD PDF

Introduction

Sheet piles and bulkheads are frequently used in waterfront construction to retain and prevent erosion of soil. In most cases, the dry/wet cycle near the splash zone, defined as the region between the low and high tide water levels, gets damaged with age.  If the seawall is made of reinforced concrete, water penetrates through the cracks in concrete, causing corrosion of the reinforcing steel.   Similarly, walls made with steel or timber get damaged and weaken with age.  In extreme cases, the section losses are large enough to allow soil to wash out through cracks and holes in the wall (Fig. 1), that could cause settlement in the ground supported by the bulkhead.  In most cases, the portion of the seawall that is embedded in soil remains relatively undamaged because the lack of oxygen at those depths cannot support the ongoing corrosion of the wall.

Following decades of research and development, the author has recently received a patent on a system for repair of such structures that is presented here (US Patent #10,968,631).

Current Repair Techniques

There is not a universally accepted method for repair of corroded seawalls.  In some cases, when there is significant loss of the steel thickness or section loss, new steel plates can be welded to the corroded wall.   However, obtaining permits for welding under water is difficult  and the cost of such welding is significant.  Furthermore, just as the original seawall corroded, the unprotected repaired steel wall is susceptible to corrosion.

Another repair technique is to clean the surface of the wall from any rust and then apply a layer of a corrosion-inhibiting paint.  The logic is that the applied coating will deprive the steel wall from exposure to oxygen and water and this will stop the corrosion process.  While this solution may seem reasonable, there are two problems with this approach.  First, this method requires the construction of a coffer dam in front of the wall and dewatering the area to give the crew access to the wall; this is a costly process, especially in deep waters.  Second, the longevity of any coating applied will remain a concern.  Once that coating begins to crack, water and oxygen will penetrate through those cracks, fueling the corrosion process. 

Sheet Pile Repair (SPiRe^®) System

The recently developed and patented system described here overcomes the above shortcomings.  The solution is called Sheet Pile Repair (SPiRe^®) system.   SPiRe^® is comprised of rigid Fiber Reinforced Polymer (FRP) panels that are custom made for each project.  The panels follow the sandwich construction technique that is commonly used in aerospace and shipbuilding industries. A proprietary lightweight core around 0.2-0.3 in. thick is sandwiched between one or more layers of unidirectional or biaxial glass or carbon fabric.  The core thickness and the type and strength of the fabric layers is a function of the project requirements and will be designed for each application.  The overall thickness of the product is typically less than 0.5 in. (Fig. 2).

The panels are made using a resin infusion process.  As shown in Fig. 2, these could be custom made to any shape to match the existing wall geometry.  However, the most common are the flat panels that are 4 ft wide.  They can be manufactured in any length and can also be cut to the desired length in the field. While the panels can be used for structural strengthening, the projects completed to date have been for providing a corrosion protective system for the walls only.

Field Installation

In the first application of the SPiRe^® system, a 926-ft long wall varying in height from 4-14 ft was repaired in Chesapeake, VA.  The corrugated metal sheet pile is part of a pulp and paper factory and was severely damaged by corrosion.  The owners were concerned about chemicals on the plant ground leaching through the wall into the river causing damage to the environment.      

The construction steps are shown in Fig. 3.  Holes are cut in the steel sheet pile and stainless-steel J-bolts are used as tie-down elements.  The J-bolts are secured to the wall using a washer and a nut.  When required, reinforcing bars can also be placed in the cavity of the wall. It is preferred to use glass FRP bars for such applications to eliminate any concern for future corrosion of these bars.  The panels have a 4-in. wide depressed lip along one of the long edges (Fig. 2) that is used for connecting the panels together.  A special underwater epoxy paste is applied to create a sealed overlapping region.  The panels are also secured to the seawall along those overlapping edges with the J-bolts at predetermined locations.  This creates an impervious stay-in-place form in front of the deteriorated seawall.  If necessary, whalers can also be temporarily placed horizontally to prevent excessive deflection of the panels during the placement of the grout. 

The bottom of the panels can penetrate by about 16-20 in. into the soft silt at the riverbed.  In other applications, a temporary seal can be produced at the lower elevation where the panels end. When placing the grout by the tremie method, about 6 to 12 inches of grout is placed along the base of the wall and allowed enough time for this grout to harden and create a seal.  The remaining height of the wall can be filled with grout in a single lift if desired.  The whalers are removed, and the ends of the J-bolts can be trimmed and if desired, covered with a cap.

The repair system which is a combination of the FRP SPiRe^® panels and the grout provides an impervious layer.  This system will prevent moisture and oxygen reaching the corroding wall.  Because oxygen is the fuel to the corrosion process, by depriving the wall of this fuel source, the corrosion process comes to near halt.  For this project, the impervious wall will also prevent leaching of any toxic materials from the plant into the river.   A video of this project is available at: www.TinyURL.com/SPiRe-WestRock

Among the advantage of this solution is that it eliminates the need for coffer dams.  Also, the lightweight panels can be easily lifted by workers, making the installation much easier and faster for hard-to-reach areas.  The repair system includes no steel components that could corrode. Once installed, the solution is expected to provide a 50+ year service life of virtually maintenance free protection for the wall.

Several ports and private piers have been repaired with this technique to date.  A major upcoming project is in the port of Melbourne, Australia where a 1 km long seawall is scheduled for repair in 2023.  The US military has also used this solution on several projects globally including at this Coast Guard facility in Tongue Point, Oregon: www.TinyURL.com/SPiRe-USCG.  This video shows another application for repair of corrosion-damaged steel bulkheads at the Port of West St. Mary, LA: www.TinyURL.com/SPiRe-StMary 

While the applications to date have been to repair and prevent further corrosion of the walls, the SPiRe^® system can be designed to contribute to the load-carrying capacity of a wall also.  Such repair options must be evaluated and designed for on a case-by-case basis.

The General Contractor for the project described in this paper was FRP Construction LLC, Tucson, AZ.  The diving subcontractor was Crofton Industries, Portsmouth VA.  The efforts of both these companies is gratefully acknowledged.

DOWNLOAD PDF