U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Back to Publication List        
Publication Number:  FHWA-HRT-17-096     Date:  October 2017
Publication Number: FHWA-HRT-17-096
Date: October 2017


Field Testing of an Ultra-High Performance Concrete Overlay




Typically, UHPC-class materials are composed of portland cement, supplementary cementitious materials (SCM) such as silica fume, fine sand, silica powder, high volumes of chemical admixtures to promote flow, and high volumes of fiber reinforcement. Gradation of solids found in UHPC is engineered to produce a dense matrix with a discontinuous pore structure, which results in exceptional durability, high stiffness, compressive strengths greater than 21.7 ksi (150 MPa), and tensile strengths higher than conventional concrete-like materials. Most UHPCs deployed in structural applications use high-strength steel microfiber reinforcement, and the percentage of fibers per unit volume of material is typically equal to or greater than two percent. This allows UHPC-class materials to undergo large tensile deformations prior to loss of tensile capacity. That is, UHPC exhibits post-cracking load-carrying capacity and tensile ductility. Finally, most UHPCs are designed to be self-consolidating and flowable under gravity without mechanical assistance.

Currently, no definition of UHPC is universally accepted, and the specific properties that define this class of materials are still being debated. FHWA’s definition does, however, provide some insight into the constituents and properties of UHPC-class materials. FHWA defines UHPC as:

a cementitious composite material composed of an optimized gradation of granular constituents, a water-to-cementitious materials ratio less than 0.25, and a high percentage of discontinuous internal fiber reinforcement. The mechanical properties of UHPC include compressive strength greater than 21.7 ksi (150 MPa) and sustained post-cracking tensile strength greater than 0.72 ksi (5 MPa). UHPC has a discontinuous pore structure that reduces liquid ingress, significantly enhancing durability compared to conventional concrete. (Graybeal 2011)

Finally, like conventional concretes, UHPC properties are highly dependent on the curing environment. UHPCs subjected to heat and/or steam during curing tend to exhibit enhanced mechanical and durability properties compared with UHPCs cast in the field; these are referred to as “field-cast” UHPCs. (Graybeal 2006; Graybeal and Stone 2012)

Table 2 lists typical properties of field-cast UHPC. The data shown in this table was collected through a series of research projects executed at the FHWA TFHRC. (Graybeal 2006; Graybeal and Stone 2012; Graybeal and Baby 2013; Swenty and Graybeal 2013)

Table 2. Typical properties of field-cast UHPC.

Material Characteristic Average Result
Density 155 lb/ft3 (2,480kg/m3)
Compressive strength (ASTM C39; 28-day strength) 24 ksi (165 MPa)
Modulus of elasticity (ASTM C469; 28-day modulus) 7,000 ksi (48 GPa)
Direct tension cracking strength (uniaxial tension with multiple cracking) 1.2 ksi (8.5 MPa)
Split cylinder cracking strength (ASTM C496) 1.3 ksi (9.0 MPa)
Prism flexural cracking strength (ASTM C1018; 12 in (305-mm)) span) 1.3 ksi (9.0 MPa)
Tensile strain capacity before crack localization and fiber debonding >0.003
Long-term creep coefficient (ASTM C512; 11.2 (77 MPa) load) 0.78
Long-term shrinkage (ASTM C157; initial reading after set) 555 microstrain
Total shrinkage (embedded vibrating wire gage) 790 microstrain
Coefficient of thermal expansion (AASHTO TP60-00) 8.2 x10-6 in/in/°F
(14.7 x10-6 mm/mm/°C
Chloride ion penetrability (ASTM C1202; 28-day test) 360 coulombs
Chloride ion penetrability (AASHTO T259; 0.5-in (12.7-mm) depth) <0.10 lb/yd3
(<0.06 kg/m3)
Scaling resistance (ASTM C672) No scaling
Abrasion resistance (ASTM C944 2x weight; ground surface) 0.026 oz. (0.73 g) lost
Freeze-thaw resistance (ASTM C666A; 600 cycles) RDM = 99 percent
Alkali-silica (ASTM C1260; 28-day test) Innocuous
RDM = Relative dynamic modulus of elasticity; ASTM = American Society of Testing and Materials;
AASHTO = American Association of State Highway and Transportation Officials.



Bridge decks are commonly rehabilitated using overlays depending on the cause of deck deterioration, available budget, and desired service life of the rehabilitated structure. Common overlay materials include conventional concrete, high-performance concretes (HPCs), latex-modified concretes (LMCs), asphalt, and polymer-based materials. There are two common methods for installing a traditional bridge deck overlays: 1) without removal of cover concrete; or 2) after full or partial removal of cover concrete. When the cover concrete remains intact, localized damage to the deck may be repaired prior to placing the overlay, and the deck may be milled or roughened to promote bonding between the deck and the overlay. When the second technique is employed, cover concrete is removed by hydrodemolition, milling or other mechanical means; in previous deployments, most UHPC overlays were placed using this second method. The performance objectives of a bridge deck overlay might include: protecting the underlying deck and reinforcement from contaminates, providing additional strength and stiffness to the deck system, or extending the service life of the overall structure. A number of properties listed in Table 2 make UHPC a viable option for use as a bridge deck overlay material. UHPC overlays offer the following potential advantages:

The primary difference between typical UHPC formulations and UHPCs that have been specially formulated for overlay applications are rheological properties. As noted above, most UHPCs are formulated to flow under the force of gravity. That is, although UHPC is viscous, it has a low fluid yield stress, which allows that material to flow freely under gravity. UHPCs for overlay applications are typically formulated to be thixotropic. Thixotropy is a time-dependent shear thinning property of a non-Newtonian fluid. A thixotropic material will remain solid-like under static conditions and with flow when agitated or sheared. This is an important property to note because bridge decks are not level. Thus, if a typical (nonthixotropic) UHPC is used as a bridge deck overlay, it will flow from the crown, or high side, of the super elevation to low points on the structure.

Figure 4 and figure 5 compare the flow table test (ASTM C1437) response of typical, nonthixotropic UHPC and thixotropic UHPC that was specially formulated for bridge deck overlays. (ASTM C1437-15 2015) Once the miniature slump cone was removed, the nonthixotropic UHPC spread across the flow table, while the thixotropic UHPC remained in solid-like state.

Figure 4. This photo shows a nonthixotropic UHPC-class material during an ASTM C1437 flow table test after removal of miniature slump cone. The photo shows that the UHPC has spread across the brass flow table. The flow diameter is approximately 8.5 in (216 mm). Figure 5. This photo shows a thixotropic UHPC-class material during an ASTM C1437 flow table test after removal of miniature slump cone. The photo shows that the UHPC has not spread far across the brass flow table. The flow diameter is approximately 4.5 in (114 mm).
Figure 4. Photo. Nonthixotropic formulation. Figure 5. Photo. Thixotropic formulation.

Other than the difference in rheological properties, the two formulations were similar in regard to mechanical and durability properties. For example, figure 6 shows a comparison of direct tension behavior of two UHPCs with nearly identical formulations, except that one was formulated to be thixotropic; the direct tension test method is shown in figure 7. (ASTM C1583/C1583M-13 2013) The materials exhibit similar elastic stiffness and post-cracking tensile ductility. Further, figure 8 shows a comparison of the same two UHPC materials and their associated bond strength to existing concrete measured according to ASTM C1583 (figure 9). (ASTM C1583/C1583M-13 2013)

The data shown reflect bond strength of UHPC bonded to an existing concrete surface that was roughened (using an exposed aggregate treatment) prior to placement of UHPC. Both UHPC materials exhibited similar bond strength despite having different rheology.

Figure 6. This line graph shows a line plot with two entries. The lefthand vertical axis represents Axial Stress in ksi and ranges from 0 ksi to 2 ksi. The righthand vertical axis represents Axial Stress in MPa and ranges from 0 ksi to 13.8 ksi. The horizontal axis represents Average Axial Strain and ranges from 0 in/in (mm/mm) to 0.01 in/in (mm/mm). One data entry (solid line) is for the thixotropic formulation, and the second (dashed line) is for the nonthixotropic form. Both formulations have approximately the same initial stiffness. The nonthixotropic formulation exhibits approximately plastic behavior at an axial stress of 1.15 ksi (7.9 MPa). The thixotropic formulation exhibits approximately plastic behavior at an axial stress of 1.4 ksi (9.6 MPa).

Figure 6. Graph. Comparison of the direct tension behavior of thixotropic and nonthixotropic UHPC.


Figure 7. This diagram shows the direct tension test method used to determine the axial stress-strain behavior of UHPC. This test was used to capture the data shown in figure 6. The figure shows a rectangular specimen being gripped at each end by a wedge-style grip. The wedge grips contact gripping plates are attached to the specimen. An extensometer is shown installed at the mid-height of the specimen. Two arrows show that the specimen is being subjected to tensile loading.

Figure 7. Illustration. Direct tension test method for UHPC.


Figure 8. This figure shows a bar chart with eight data entries. The lefthand vertical axis represents Peak Tensile Stress at Failure in ksi and ranges from 0 ksi to 1.4 ksi in 0.2 increments. The righthand vertical axis represents Peak Tensile Stress at failure in MPa and ranges from 0 ksi to 9.6 ksi. The horizontal axis depicts the direct tension strength response of nonthixotropic and thixotropic UHPCs at 7 and 14 days of age. The plot depicts both the interface bond strength and the direct tensile strength of UHPC. The data shown in the plot from left-to-right are as follows: Nonthixotropic UHPC at 7 days of age had an interface bond strength of 0.37 ksi (2.4 MPa) and a direct tension strength of 0.92 ksi (6.3 MPa). Thixotropic UHPC at 7 days of age had an interface bond strength of 0.32 ksi (2.2 MPa) and a direct tension strength of 0.8 ksi (5.5 MPa). Nonthixotropic UHPC at 14 days of age had an interface bond strength of 0.41 ksi (2.8 MPa) and a direct tension strength of 0.95 ksi (6.5 MPa). Thixotropic UHPC at 14 days of age had an interface bond strength of 0.36 ksi (2.5 MPa) and a direct tension strength of 0.92 ksi (6.3 MPa).

Figure 8. Graph. Comparison of the direct tension bond strength of thixotropic (Thix) and nonthixotropic (Non-Thix) UHPC formulations.


Figure 9. This illustration shows the ASTM C1583 test method setup, where a UHPC topping is cast over a precast concrete base slab, thus creating an interface for bond evaluation. The illustration shows the loading setup of the specimen, which consists of a single tensile load applied on a steel disc previously glued on the top surface of the UHPC topping. The test specimen is formed by partially drilling a core perpendicular to the surface and penetrating down to the concrete material (approximately 1 in (25.4 mm) below the UHPC-concrete interface).

Figure 9. Illustration. ASTM C1583—Direct tension bond pull-off test method.


The cost of UHPC is generally higher than most highway bridge construction materials. Further, the material cost and bid line-item cost for UHPC-class materials can differ substantially. The Federal Highway Administration (FHWA) provides high-quality information to serve Government, industry, and the public in a manner that promotes public understanding. Standards and policies are used to ensure and maximize the quality, objectivity, utility, and integrity of its information. FHWA periodically reviews quality issues and adjusts its programs and processes to ensure continuous quality improvementcompares the approximate cost of various overlay solutions. The UHPC costs shown reflect only material costs but not associated installation costs and are thus relatively low. In general, the cost of UHPC overlays in the United States will likely be higher than most traditional overlay solutions until the technology becomes established. In Switzerland, UHPC overlays are commonly deploy, and, therefore, costs are becoming competitive.

Table 3 lists the approximate cost ranges of a UHPC on a signature bridge structure, the Chillon Viaduct, in Switzerland; this cost is comparable to the upper-end cost of some conventional solutions.

Table 3. Approximate cost ranges of bridge deck overlay solutions.

Overlay Type Overlay Thickness—in (mm) Cost—$/ft2 ($/m2)
High-performance concrete* 1–5 (25–127) 17–25 (183–269)
Low slump concrete* 1.5–4 (38–102) 13–19 (140–204)
Latex-modified concrete* 1–5 (25–127) 18–39 (193–419)
Asphalt with a Membrane* 1.5–4 (38–102) 3–8 (32–86)
Polymer-based* 0.13–6 (3–152) 10 –17 (107–183)
Nonproprietary UHPC 1–2 (25–52) 3–6 (32–64)†
Proprietary UHPC 9–18 (97–184)
Rehabilitation of the Chillon Viaduct (Switzerland)
Proprietary UHPC Overlay 1.6 (40)** 20 (215)**
Other Rehabilitation Solutions
Deck Replacement* 43 - 53 (462–570)
*Data collected from Krauss et al. (2009); The costs shown reflect average values from low and high ranges.
**Data collected from Brühwiler et al. (2015); Price reflects cost of material and installation.
†Price reflects material cost only; Assumes UHPC CY cost of $1,000.
‡Price reflects material cost only; Assumes UHPC CY cost of $3,000.
—Not applicable.



The concept of using UHPC as an overlay was pioneered in Switzerland and has since been deployed on numerous bridges in Europe. A number of laboratory studies were conducted prior to field deployments. Studies have been conducted on the flexural and combine flexure-shear behvior of reinforced concrete (RC) members with thin UHPC overlays. (Habel, Denarié, and Brühwiler 2007; Noshiravani and Bruhwiler 2013)

Findings indicated that applying a thin layer of UHPC, 1.25- to 4-in (32- to 102-mm) thick, to an existing reinforced concrete member to form a composite UHPC-RC member can increase stiffness, decrease crack width and spacing, and increase load-carrying capacity of the element. Further, results indicated that UHPC-RC elements can achieve composite behavior if the concrete surface was prepared using hydrodemolition prior to casting UHPC.

A study conducted at Iowa State University further investigated the behavior of reinforced concrete slab-like element strengthened using UHPC overlays. (Aaleti, Sritharan, and Abu-Hawash 2013) Sixty slant shear bond tests and three large-scale flexural tests were conducted. Slant shear tests were used to investigate the effect of existing concrete surface roughness on the bond between UHPC and concrete. Subsequently, large-scale flexural tests were used to investigate the system behvior of the RC member with the UHPC overlay. It was concluded that increasing the surface roughness of the existing concrete increased the bond strength between concrete and the overlay material. It was also concluded that a minimum surface roughness of 0.08-in (2 mm) was required for a reinforced concrete member with a UHPC overlay to achieve good flexural performance.

A study conducted at Michigan Technological University investigated using UHPC as a bridge deck overlay material considering time-dependent effects such as shrinkage and stresses within the system during service. (Shann 2012) A series of restrained ring shrinkage tests were conducted according to a modified version of the American Association of State Highway and Transportation Officials (AASHTO) PP34-99 test method. A parametric study was executed using two-dimensional and three-dimentional finite-element analysis (FEA) to investigate the behavior of a UHPC overlay in bridges with varied girder spacing, overlay thicknesses, support conditions, and bridge deck thicknesses. The influence of relative stiffness between the UHPC overlay and the underlying deck concrete was investigated. A series of HS-20 design truck loading configurations were also considered.

The debonding and interfacial shear stress was found to increase with overlay thickness, while tension stress that would produce cracking or fracture in a UHPC overlay decreased with increased overlay thickness. These effects were largely attributed to shifts in the neutral axis location. The authors concluded that, when effects of restrained shrinkage and live load were combined, the tensile stresses in a UHPC overlay controlled the design. The authors also concluded that UHPC overlays may not be compatible on bridge decks thicker than 10 in (254 mm) due to neutral axis shifting, which causes large tensile stresses in the UHPC overlay.

These results were obtained from a limited FEA study. The majority of results were obtained using a conservative, overly simplified plate model. When an actual bridge structure with a UHPC overlay was analyzed, the interfacial shear and debonding stresses at the UHPC-concrete interface were relatively low.

Researchers have also investigated the use of UHPC as a topping layer on orthotropic steel bridge decks. Research conducted by Toutlemonde et al. studied the feasibility of using a thin layer of UHPC to reduce the potential of fatigue-induced damage and cracking in orthotropic steel decks. (Toutlemonde et al. 2013) Two systems were tested: (1) where a thin layer of UHPC was cast over wire mesh placed atop an orthotropic steel deck specimen; and (2) where a thin layer of UHPC was cast directly onto an orthotropic steel deck specimen that employed short shear studs to improve composite action between the deck and the UHPC overlay. Results showed that imperfect bonding between the steel deck plate and the UHPC topping caused slippage and some cracking at the interface under extremely high loading. Nevertheless, experimental and numerical modeling found that using a UHPC overlay could reduce stresses in the orthotropic steel deck compared to a deck with a bituminous overlay.

Since 2004, UHPC overlays have been deployed on more than 20 bridges in Switzerland and on one bridge in Slovenia. (Brühwiler and Denarié 2013; Sajna, Denarié, and Bras 2012) Applications include waterproofing bridge decks after widening projects, simultaneous waterproofing and strengthening, and rehabilitation of crash barriers. A detailed summary of these applications is provided by Brühwiler and Denarié. (Brühwiler and Denarié 2013)

The most notable UHPC overlay deployment was in the rehabilitation of the Chillon Viaduct. This 1.34 mi (2.15 km) long elevated bridge structure runs along Lake Geneva in the Swiss canton of Vaud. The viaduct is composed of two parallel post-tensioned concrete box girder structures built in the late 1960s. In 2012, a significant rehabilitation project began, during which inspectors noted that a reinforced concrete deck was exhibiting the initial signs of alkali-aggregate reactivity (AAR). A UHPC overlay was selected to act as both a thin waterproofing layer and a strengthening mechanism. The existing deck was relatively thin—7.1 in (180 mm)—and required a structural strengthening solution that did not cause a significant increase in dead load. A photo of the structure is shown in figure 10, and a typical section is shown in figure 11.

Figure 10. This figure shows a photo of the parallel post-tensioned box girder structures that compose the Chillon Viaduct.

Figure 10. Photo. Chillon Viaduct: Parallel post-tensioned box girder structures.


Figure 11. Typical section of a single post-tensioned box girder. This figure is an illustration of a typical section of a single post-tensioned box girder. The typical section shows that the existing superstructure is partially hollow. It also shows the thin UHPC overlay that was installed on the existing structure.

Figure 11. Illustration. Chillon Viaduct Details: Typical section of a single post-tensioned box girder.


Figure 12 and figure 13 show photos taken during construction of the Chillon Viaduct UHPC overlay project. The deck rehabilitation was done in two phases completed in the summers of 2014 and 2015—one for each of the two parallel viaduct structures. The UHPC deck overlay for each structure was completed in less than 30 working days. The UHPC overlay, which had a thixotropic formulation, was installed using a specially designed placement machine (shown in figure 13) after completing the UHPC overlay, a 3.2-in (80-mm) asphalt layer was installed as a ride surface.

Figure 12. This figure shows a photo of the Chillan Viaduct UHPC overlay project during the second construction phase. The photo shows two roadway surfaces: the lefthand roadway depicts what the roadway looks like after the overlay has been installed along with the ride surface. Vehicles are shown driving on this completed portion; the righthand roadway shows construction operations during UHPC overlay installation.

Figure 12. Photo. Arial view of the Chillon Viaduct UHPC overlay construction project.


Figure 13. This figure shows a photo of the UHPC placement machine actively placing the UHPC overlay on a prepared section of bridge deck. The prepared bridge deck section has been formed, and a layer of steel reinforcing bars has been placed.

Figure 13. Photo. Placement of UHPC in the Chillon Viaduct project.




Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101