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
Publication Number: FHWA-HRT-06-103
Date: August 2006

Material Property Characterization of Ultra-High Performance Concrete



UHPC is a new type of concrete that exhibits properties of enhanced strength, durability, and long-term stability. The objective of this research was to evaluate the material characteristics of UHPC for potential use in highway bridge applications.

The experimental phase of this research focused on determining the mechanical and durability behaviors of UHPC. More than 1,000 individual specimens were tested to determine the material characteristics of UHPC. The tests determined the compressive and tensile behaviors, the longterm stability, and the durability of UHPC. The analytical phase of this research combined, analyzed, and elaborated on the results from the experimental phase. This phase included developing predictor equations for some basic properties of UHPC.

The conclusions of this study are presented in section 5.2. A brief discussion of ongoing and potential future research topics follows in section 5.3.


The following conclusions are based on the research presented in this report.

  1. UHPC displays significantly enhanced material properties compared with normal and HPC.
  2. Steam-based treatment of UHPC tends to significantly enhance its material properties. Three steam—based treatments—steam, delayed steam, and tempered steam–were investigated and compared with UHPC that was not subjected to a curing treatment after casting. In general terms, steam treatment increases UHPC’s compressive strength by 53 percent to 193 MPa (28 ksi), increases its modulus of elasticity by 23 percent to 52.4 GPa (7,600 ksi), decreases its creep coefficient from 0.78 to 0.29, and virtually eliminates long-term shrinkage. Steam treatment also decreases chloride ion penetrability to a negligible level and significantly enhances abrasion resistance. The enhancements of material properties affected by the delayed steam and tempered steam treatments are similar to those of the steam treatment but of a slightly lesser magnitude.
  3. UHPC exhibits very high compressive strengths, regardless of the curing treatment applied. The average 28-day compressive strengths of steam, delayed steam, tempered steam, and untreated UHPC were found to be 193, 171, 171, and 126 MPa (28.0, 24.8, 24.8, and 18.3 ksi). The compressive strength of steam–treated UHPC was found to have stabilized by the completion of the curing treatment. Thus, steam-treated UHPC can reach its full compressive strength within 4 days after casting.
  4. The mixing time and rheological properties of fresh UHPC are influenced by the concrete mixer design, the ambient environmental conditions in the mixer, and the elapsed time since blending of the premix. A 1934 vintage pan mixer was used successfully to mix the UHPC; however, the inability of this mixer to impart significant energy into the mix resulted in extended mixing times. Low humidity within the mixer and in the mix room can result in stiffer UHPC. Older UHPC premix requires more mixing to achieve the correct rheological properties, likely due to the agglomeration of fine particles in the premix during storage.
  5. The set time of UHPC is significantly delayed compared with normal concrete; final set does not occur until 12 to 24 hours after casting. This time to set could also be longer depending on the admixtures and on other constituents in the mix.
  6. Once setting has initiated, UHPC gains compressive strength very rapidly. If maintained at normal laboratory temperatures, UHPC compressive strength will increase to over 70 MPa (10 ksi) by 2 days after setting. Subsequently, the rate of strength gain will decrease; 97 MPa (14 ksi) will be reached by 10 days after setting.
  7. The compressive strength of UHPC is not affected by the specimen geometry used to determine the result. Cylinders with 51-, 76-, and 102-mm (2-, 3-, and 4-inch) diameters were tested according to ASTM C39, and 51-mm (2-inch) and 100-mm (4-inch) cubes were tested according to ASTM C109. The testing was conducted for two batches of steam-treated UHPC and for one batch of untreated UHPC. In all cases, the compressive strength results did not vary by more than 8 percent from the 76-mm (3-inch) cylinder control result. However, the 51-mm (2-inch) cubes and cylinders did tend to exhibit a larger standard deviation. The minimal preparation requirements for a cube specimen and the similarity of results mentioned above make the 100-mm (4-inch) cube a viable specimen geometry for UHPC compressive strength determination.
  8. The curing conditions present during and just after the setting of UHPC can significantly affect the final properties of the concrete. In the untreated case, concrete cylinders that were stripped as final set was being reached exhibited 25 percent lower 28-day compressive strengths compared with those stripped 1 day after setting was complete. A 30-percent difference was observed in the steam-treated case. These strength differences are likely due to the relatively more permeable nature of UHPC at earlier ages combined with the very low moisture content in the UHPC. This low moisture content results in a loss of water to the surrounding atmosphere and thus reduced hydration of the concrete.
  9. The tensile strength of UHPC, both before and after tensile cracking, is significantly higher than the strength that occurs in normal concrete. Four test methods were implemented to capture the tensile strength of UHPC. The combined results of these tests indicate that the tensile cracking strength of UHPC is approximately 9.0 MPa (1.3 ksi) after the steam-based curing treatment and approximately 6.2 MPa (0.9 ksi) without any treatment. Qualitatively, UHPC was observed to exhibit similar levels of tensile strength after cracking; however, in general, these tests were unable to indicate specific postcracking strengths.
  10. The ASTM C496 split-cylinder tension test, modified to capture first cracking, provides the clearest indication of the tensile cracking strength of UHPC. Through the use of a hydraulically controlled test machine and minimal instrumentation, the tensile cracking strength of UHPC can be obtained. Other tensile tests require more extensive instrumentation, specialization of specimen geometry, and more sophisticated loading equipment. Unfortunately, the split-cylinder test is not useful for determining the postcracking tensile stress capacity, because the biaxial state of stress applied to the cylinder does not accurately mimic the conditions that are normally associated with tensile regions in a structural member.
  11. The modulus of rupture defined by the ASTM C1018 prism flexure test overestimates the tensile cracking strength of UHPC by approximately 60 percent. This result was confirmed through the completion of this test on 51-mm (2-inch) square cross sections with 152-, 229-, 305-, and 381-mm (6-, 9-, 12-, and 15-inch) spans, and on 76- by 102-mm (3- by 4-inch) cross sections with a 305-mm (12-inch) span.
  12. The ASTM C1018 prism flexure test provides a clear means of comparing the postcracking tensile behavior of various fiber-reinforced concretes. UHPC, regardless of curing treatment, performed exceptionally well according to the toughness indices defined by this test. For example, the I20 toughness index normally ranges from 1 to 25 for fiber-reinforced concretes. In UHPC, the I20 results ranged from 28 to 32. Although these results cannot be directly reinterpreted to apply to full-scale structural members, they do indicate that UHPC can continue to carry significant tensile loads after cracking.
  13. UHPC displays durability properties that are significantly beyond those normally associated with concrete. Regardless of the curing treatment, the ASTM C666 relative dynamic modulus was at least 95 percent of the original value after more than 600 freeze-thaw cycles. UHPC exhibited no scaling under the ASTM C672 test, even after undergoing approximately 200 cycles. The chloride ion penetrability as measured by ASTM C1202 was below 50 coulombs for UHPC that had undergone the steam-based treatment, and was 360 coulombs for untreated UHPC 28 days after casting. The untreated UHPC results dropped to 76 coulombs by 56 days after casting. UHPC was found to be innocuous to alkali-silica reaction.
  14. Exposing cracked UHPC split cylinders to an aggressive environment did not result in any noticeable decrease in the peak tensile load-carrying capacity. Tight cracks, as might be observed in a highly stressed tensile flexural region of an I-girder, were created by loading cylinders in a split-cylinder configuration. These cracks were on the order of 0.005 mm (0.0002 inch) wide. A cracked face of the cylinder was then ponded with a sodium chloride solution as specified in AASHTO T259. After 90 days of ponding, the cylinders were tested for peak split–cylinder tensile strength. The peak load-carrying capacity of either steam-treated or untreated UHPC did not have a discernable decrease after cracking, thus indicating that the sodium chloride solution did not enter the cracks and did not cause the fiber reinforcement to deteriorate.
  15. UHPC exhibits shrinkage behaviors that are somewhat different from those of normal concrete. In total, UHPC tends to exhibit approximately 800 microstrain of shrinkage as measured from casting through 1 year. However, shrinkage initiation is affected by the delayed set times associated with UHPC, and the majority of the shrinkage occurs in a short time frame just after the concrete has set. Unrestrained shrinkage rates of over 60 microstrain per hour were observed during the period of rapid strength gain just after setting. Without any curing treatment, UHPC will continue to shrink at an ever-decreasing rate. Steam treatment accelerates the shrinkage to such an extent that the entirety of the shrinkage occurs during the 2-day treatment, and the UHPC is then stabilized against further shrinkage. Also, the total shrinkage in steam-treated UHPC tends to be slightly higher than the asymptotic shrinkage approached by untreated UHPC.
  16. Large compressive stresses on relatively low-strength UHPC can cause significant short-term creep. This situation is akin to the stressing of prestressed girders. Eight to 13 ksi compressive strength UHPC was loaded to compressive stresses between 60 percent and 90 percent of the strength. During the 30 minutes following the load application, the UHPC exhibited 30-minute creep coefficients between 0.32 and 0.85. UHPC loaded to over 90 percent of its compressive strength failed under the sustained load. The creep that occurred over this short load duration indicates that the total long–term creep of UHPC loaded at this compressive strength would be much higher than that observed in the long-term creep testing.


The findings from this report suggest a number of potential topics for future research:

  1. Develop optimized bridge girders that take advantage of the material properties of UHPC. These bridge girders should use the tensile and compressive capacities of UHPC, while also enhancing the design life of the bridge as a whole by eliminating many of the less durable components of a normal bridge.
  2. Fabricate full-scale, optimized UHPC bridge girders to resolve problems associated with casting slender concrete members with fiber-reinforced concrete.
  3. Develop a practical test to quantitatively determine the postcracking uniaxial tensile behavior of UHPC.

The research program discussed herein has already been extended to encompass a portion of the topics listed above.

Previous    Table of Contents    Next


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