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

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

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-13-060    Date:  June 2013
Publication Number: FHWA-HRT-13-060
Date: June 2013


Ultra-High Performance Concrete: A State-Of-The-Art Report for The Bridge Community



UHPC formulations often consist of a combination of portland cement, fine sand, silica fume, high-range water-reducing admixture (HRWR), fibers (usually steel), and water. Small aggregates are sometimes used, as well as a variety of chemical admixtures. Different combinations of these materials may be used, depending on the application and supplier. Some of these are described in this section.

The UHPC used most often in North America for both research and applications is a commercial product known as Ductal®. Table 1 shows a typical composition of this material.(22)

Table 1. Typical composition of Ductal®

Material lb/yd3 kg/m3 Percentage by Weight
Portland Cement 1,200 712 28.5
Fine Sand 1,720 1,020 40.8
Silica Fume 390 231 9.3
Ground Quartz 355 211 8.4
HRWR 51.8 30.7 1.2
Accelerator 50.5 30.0 1.2
Steel Fibers 263 156 6.2
Water 184 109 4.4

Aarup reported that CRC, developed by Aalborg Portland in 1986, consisted of large quantities of steel fibers (2 to 6 percent by volume), large quantities of silica fume, and a water-binder ratio of 0.16 or lower.(23)

The following recommendations for mix proportions were developed for use with commercially available constituent materials:(24)

By optimizing the cementitious matrix for compressive strength, packing density, and flowability; using very high strength, fine-diameter steel fibers; and tailoring the mechanical bond between the steel fiber and cement matrix, 28-day compressive strengths in excess of 30 ksi (200 MPa) on 2-inch (50-mm) cubes were achieved with no heat or pressure curing.(25) In addition, a tensile strength of 5.0 ksi (34.6 MPa) at a strain of 0.46 percent was obtained. The UHPC incorporated materials available in the United States and was mixed in a conventional concrete mixer. Table 2 gives one mix proportion.

Table 2. UHPC mix proportions of CRC by weight(25)

Material Proportions
Portland Cement 1.0
Fine Sand1 0.92
Silica Fume 0.25
Glass Powder 0.25
HRWR 0.0108
Steel Fibers 0.22 to 0.31
Water 0.18 to 0.20
1Maximum size of 0.008 inches (0.2 mm)

Habel et al. reported that it is possible to produce self-consolidating UHPC for use in precast products and cast-in-place (CIP) applications without requiring heat or pressure treatment during curing.(26) This mix design was further developed and implemented in a research program conducted by Kazemi and Lubell.(27)

Holschemacher and Weißl investigated different mix proportions to minimize material costs without sacrificing the beneficial properties of UHPC. (28) Through careful selection of aggregates, cement type, cementitious materials, inert filler, and HRWR, it was possible to produce UHPC with good workability and moderate material costs.

The concept of combining different size molecular admixtures to facilitate UHPC dispersion was studied by Plank et al.(29)

The possibility of replacing silica fume in UHPC with metakaolin, pulverized fly ash, limestone microfiller, siliceous microfiller, micronized phonolith, or rice husk ash has been investigated.(30,31) The use of local materials rather than proprietary products has also been pursued.(32,33)

Schmidt et al. reported two mix proportions for a bridge in Germany.(34) The first mix contained 1,854 lb/yd 3 (1,100 kg/m3) of cement, 26-percent silica fume as a percentage of the cement content, quartz sand, 6 percent steel fibers by volume, HRWR, and a water-binder ratio of 0.14. The second mix contained 2,422 lb/yd3 (1,437 kg/m3) of cement and 9-percent steel wool and steel fibers combined.

Collepardi et al. reported that the replacement of fine ground quartz sand with an equal volume of well-graded natural aggregate with a maximum size of 0.3 inches (8 mm) did not change the compressive strength at the same water-cement ratio.(35)

Coppola et al. investigated the influence of high-range water-reducing admixture type on the compressive strength. They reported that acrylic polymeradmixtures allowed the use of lower water-cement ratios and resulted in higher compressive strengths compared with naphthalene and melamine admixtures. (36)

In a study of the durability of UHPC, Teichmann and Schmidt used the mix proportions shown in table 3.(37) Mix 1 had a maximum aggregate size of 0.32 inches (8 mm) provided by the sand. Mix 2 had a maximum aggregate size of 0.32 inches (8 mm) provided by the basalt.

Table 3. UHPC mix proportions from Teichmann and Schmidt(37)

Material Mix 1 Mix 2
lb/yd3 kg/m3 lb/yd3 kg/m3
Cement 1,235 733 978 580
Silica Powder 388 230 298 177
Fine Quartz 1 308 183 503 131
Fine Quartz 2 0 0 848 325
HRWR 55.5 32.9 56.2 33.4
Sand 1,699 1,008 597 354
Basalt 0 0 1,198 711
Steel Fibers 327 194 324 192
Water 271 161 238 141
Water-Binder Ratio 0.19 0.19 0.21 0.21

Researchers at the U.S. Army Corps of Engineers Engineer Research and Development Center have reported on a UHPC-class material referred to as Cor-Tuf. (38,39) The proportions of this UHPC are presented in table 4.

Table 4. UHPC mix proportions of Cor-Tuf by weight(38,39)

Material Proportions
Portland Cement 1.0
Sand 0.967
Silica Flour 0.277
Silica Fume 0.389
HRWR 0.0171
Steel Fibers 0.310
Water 0.208

Researchers led by Rossi at the Laboratoire Central des Ponts et Chaussees (LCPC) in Paris developed a UHPC-class material referred to as CEMTEC multiscale.(40) The proportions of this UHPC are presented in table 5.

Table 5. UHPC mix proportions for CEMTECmultiscale(40)

Material lb/yd3 kg/m3
Portland Cement 1,770 1,050
Sand 866 514
Silica Fume 451 268
HRWR 74 44
Steel Fibers 1,446 858
Water 303 180


Graybeal has summarized the mixing of UHPC as follows:

Nearly any conventional concrete mixer will mix UHPC. However, it must be recognized that UHPC requires increased energy input compared to conventional concrete, so mixing time will be increased. This increased energy input, in combination with the reduced or eliminated coarse aggregate and low water content, necessitates the use of modified procedures to ensure that the UHPC does mot overheat during mixing. This concern can be addressed through the use of a high-energy mixer or by lowering the temperatures of the constituents and partially or fully replacing the mix water with ice. These procedures have allowed UHPC to be mixed in conventional pan and drum mixers, including ready-mix trucks. (p. 2)(1)

Mixing times for UHPC range from 7 to 18 minutes, which are much longer than those of conventional concretes.(41,42) This impedes continuous production processes and reduces the capacity of concrete plants. Mixing time can be reduced by optimizing the particle size distribution, replacing cement and quartz flower by silica fume, matching the type of HRWR and cement, and increasing the speed of the mixer.(42) The mixing time can also be reduced by dividing the mixing process into two stages. High-speed mixing for 40 seconds is followed by low-speed mixing for 70 seconds, for a total time of about 2 minutes.(41)

The method of placing UHPC has an influence on the orientation and dispersion of the fibers.(43) The orientation did not affect the first cracking load but had an effect of up to 50 percent on the ultimate tensile strength in bending. The highest strengths were achieved when placement was made in the direction of the measured tensile strength. Stiel et al. reported significant differences between horizontally and vertically cast beams when tested in three-point bending.(44) The fibers in the vertically cast beams were aligned in layers normal to the casting direction. As a result, the splitting and flexural strengths were only 24 and 34 percent of the corresponding values for the horizontally cast beams. However, in a 39-inch (1-m)-thick slab, the fibers were arranged randomly. The orientation of the fibers did not have a significant effect on the compressive strength and modulus of elasticity.

Graybeal has summarized the placement of UHPC as follows:

The placement of UHPC may immediately follow mixing or be delayed while additional mixes are completed. Although the dwell time prior to the initiation of the cement hydration reactions can be influenced by factors such as temperature and chemical accelerators, it frequently requires multiple hours before UHPC will begin to set. During extended dwell time, the UHPC should not be allowed to self-desiccate.

Casting of fiber-reinforced concretes requires special considerations in terms of placement operations. UHPCs tend to exhibit rheological behaviors similar to conventional self-consolidating concretes, thus possibly necessitating additional form preparation but also allowing for reduced during-cast efforts. Internal vibration of UHPC is not recommended due to fiber reinforcement, but limited external form vibration can be engaged as a means to facilitate the release of entrapped air. (p. 3)(1)

For the UHPC beams used on the Route 624 bridge over Cat Point Creek in Richmond, VA, the contractor was required to use a plant that was prequalified for UHPC production, and a representative from the UHPC producer was required to be present.(45) The UHPC was mixed in 4-yd3 (3-m3) batches in an 8-yd3 (6-m3) twin shaft mixer and discharged into a ready-mixed concrete truck for delivery. About 20 to 25 minutes were required to load the mix, mix the UHPC, and discharge the mixer.

During discharge from the truck, cement balls were observed in the mix. This was attributed to exposure of the bags to moisture during storage. The mix was discharged into one end of the beam and allowed to flow. Only limited external vibration was applied for 1 or 2 seconds.


Curing of UHPC considers two distinct components, specifically temperature and moisture. As with any cementitious composite material, maintaining an appropriate temperature is critical to achieving the desired rate for the cementitious reactions. In addition, given the low water content in UHPC, eliminating loss of internal water by sealing the system or maintaining a high humidity environment is also critical.

The curing of UHPC occurs in two phases.(1,46) Given that UHPC tends to exhibit a dormant period prior to initial setting, the initial curing phase consists of maintaining an appropriate temperature while precluding moisture loss until setting has occurred and rapid mechanical property growth is occurring. The second curing phase may or may not include elevated temperature conditions and a high moisture environment, depending on whether accelerated attainment of particular material characteristics is desired.

Graybeal reported on an extensive program to determine material properties of UHPC using four different post-set curing procedures.(22) These involved steam curing at 194 °F (90 °C) or 140 °F (60 °C) for 48 hours, starting about 24 hours after casting; steam curing at 194 °F (90 °C), starting after 15 days of standard curing; and curing at standard laboratory temperatures until test age.

These three steam-curing methods increased the measured compressive strengths and modulus of elastic, decreased creep, virtually eliminated drying shrinkage, decreased chloride ion penetrability, and increased abrasion resistance. The enhancements achieved by the lower steam temperature and delayed steam curing were slightly less than achieved by steam curing at the higher temperature. The specimens steam cured at 194 °F (90 °C) after 24 hours reached their full compressive strengths within 4 days after casting. Chapter 3 of this report presents more details of the test results.

More recent work by Graybeal has focused on characterizing the performance of ambient-cured UHPC.(47) This research stems from the recognition that accelerated curing in a steam environment is frequently not practical and also that the ambient-cured properties of UHPC are appropriate for many applications.

Ay compared the compressive strength of 4-inch (100-mm) cubes cured by the following three methods:(48)

The UHPC cubes stored in water followed by air curing had slightly higher compressive strengths than cubes cured by the other two methods.

The compressive strength of UHPC can be increased considerably by using post-set heat curing.(49) Heinz and Ludwig showed that the heat curing at various temperatures between 149 and 356 °F (65 and 180 °C) produced 28-day compressive strengths as high as 41 ksi (280 MPa) compared with strengths of 25 and 27 ksi (178 and 189 MPa) when cured at 68 °F (20 °C). Higher curing temperatures resulted in higher compressive strengths. In addition, the strengths at the end of the curing period at about 48 hours after casting were about the same as the corresponding 28-daystrengths. The authors also concluded that curing at 194 °F (90 °C) presented no danger of delayed ettringite formation.(49)

Schachinger et al. observed that initial curing at 68 °F (20 °C) for 5 days, followed by heat curing at 122 to 149 °F (50 to 65 °C), was the most favorable combination to achieve high strengths at ages up to 28 days.(50) Compressive strengths in the range of 36 to 43.5 ksi (250 to 300 MPa) were achieved at ages of 6 to 8 years.

Heinz et al. achieved compressive strengths higher than 29 ksi (200 MPa) at an age of 24 hours after 8 hours storage at 68 °F (20 °C) followed by 8 hours at 194 °F (90 °C) in water.(51) Longer periods of initial storage or heat treatment resulted in higher strengths when ground-granulated blast-furnace slag was included in the UHPC. The authors obtained the highest strengths by including fly ash and autoclaving the UHPC for 8 hours at 300 °F (150 °C).

Massidda et al. showed that autoclaving at a temperature of 356 °F (180 °C) and 145 psi (1 MPa) with saturated steam produced higher compressive strengths and flexural strengths compared with specimens cured at 68 °F (20 °C).(52)


Quality control tests for UHPC in the United States have generally used the same or similar tests as those used for conventional concrete or mortar with or without modifications. Both fresh and hardened concrete properties are measured.

The flow of UHPC is frequently measured using ASTM C1437—Standard Test Method for Flow of Hydraulic Cement Mortar.(1,53) This test method is intended for use with mortars exhibiting plastic to flowable behavior, and thus it is frequently appropriate for fresh UHPC. In this test, both initial flow and dynamic flow are measured. The test is completed immediately after mixing to assess consistency among mixes and appropriateness for casting.(1) On the Route 24 bridge over Cat Point Creek, a minimum dynamic flow of 9 inches (230 mm) was sought for satisfactory workability.(45)

As different versions of UHPC are developed for different applications, alternate workability tests will be needed. For stiffer, non-self-consolidating UHPC, the ASTM C143—Standard Test Method for Slump of Hydraulic-Cement Concrete may be appropriate.(54) Scheffler andSchmidt have reported that development of stiff UHPC formulations for applications such as pavement whitetopping is feasible.(55)

The initial and final setting times of UHPC can be longer than those observed for many conventional cementitious materials. The set times are heavily influenced by the curing temperature.(47) Graybeal measured initial setting times ranging from 70 minutes to 15 hours for different UHPC formulations using the American Association of State Highway and Transportation Officials (AASHTO) T 197 test method for penetration resistance.(22,56,57) The corresponding final setting times ranged from 5 to 20 hours.

Compressive strength testing of UHPC is frequently completed using a modified version of ASTM C39—Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens.(58) The test method is modified to include an increased load rate of 150 psi/second (1 MPa/second) in response to the high compressive strength that UHPC exhibits.(47) Appropriate cylinder end preparation is critical because non-flat or non-parallel end surfaces can cause a reduction in observed compressive strength.(1) End surface preparation for cylinders with early age compressive strengths below 12 ksi can be completed using multiple methods, including capping according to ASTM C617.(1,47,59) Higher strength cylinders should have their ends ground to within 0.5 degrees.(58)

Smaller cylinders have been shown to provide strengths equivalent to traditional size cylinders. Graybeal reported that 3- by 6-inch (76- by 152-mm) cylinders exhibited similar strengths to 4- by 8-inch (102- by 203-mm) cylinders while allowing for the use of a significantly reduced testing machine capacity.(22,60) Use of 2- by 4-inch (51- by 102-mm) cylinders was not recommended because of the increased dispersion present in the results.

Research has demonstrated that the ASTM C109—Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-inch (50-mm) Cube Specimens) can also be applied to UHPC.(61) Graybeal reported that 2-inch, 2.8-inch, and 4-inch cubes exhibited compressive strengths up to approximately 7 percent greater than those observed from 3- by 6-inch and 4- by 8-inch (76- by 152-mm and 102- by 203-mm)cylinders.(22,60) Similar findings were reported by Alhborn and Kollmorgen.(62)

On the U.S. Route 6 bridge over Keg Creek in Pottawatomie County, IA, UHPC was used in the longitudinal and transverse joints between the concrete deck panels.(63) The Special Provisions for the project required the contractor to cast twelve 3- by 6-inch (75- by 150-mm) cylinders for verification of concrete compressive strength.(64) Three cylinders were to be tested to verify 10.0 ksi (69 MPa) at 96 hours, three to verify 15.0 ksi (103 MPa) for opening the bridge to traffic, and three at 28 days. The remaining three specimens were treated as reserves. Specimens were required to have their ends ground to 1 degree planeness.

For field-cast UHPC joints, the New York State Department of Transportation (NYSDOT) also requires the casting of twelve 3- by 6-inch (75- by 150-mm) cylinders for testing in sets of three.(65) One set is tested at 4 days, one set at 28 days, one set is to be supplied to the NYSDOT, and one set is treated as reserve.

For qualification testing of the proposed UHPC mix, NYSDOT requires that a minimum of sixty-four 2-inch (50-mm) cubes be cast. Testing ages are 4, 7, 14, and 28 days. Minimum compressive strengths of 14.3 ksi (100 MPa) at 4 days and 21.8 ksi (150 MPa) at 28 days are required.

Frölich and Schmidt investigated the repeatability and reproducibility of tests methods for fresh UHPC.(66) They observed that the values of the measured fresh properties were influenced by the time of measurement, mixing equipment, laboratory conditions, operator, and air-void content. The authors concluded that quality control tests should be made 30 minutes after the start of mixing and that flowable consistency should be measured using the slump flow test.


The constituent materials of UHPC generally consist of portland cement, fine sand, ground quartz, HRWR, accelerating admixture, steel fibers, and water. As a class, UHPCs have high cementitious materials contents and very low water-cementitious materials ratios. UHPC can be mixed in conventional mixers but the UHPC mixing time is longer than for conventional concrete. The method of placing UHPC has an influence on the orientation and dispersion of the fibers, which influences the tensile properties of the UHPC. The properties of UHPC are affected by the method, duration, and type of curing. As with conventional concrete, heat curing accelerates the development of strength and related properties. Delaying the application of heat for several days can enhance the measured properties, although it may not be compatible with the rapid production in precasting operations. Smaller size cylinders have been used in quality control for measurement of compressive strengths.