Evaluating CEN/TS 12390-9 To Assess Concrete for UK Freeze-Thaw Exposure Conditions

  • Christopher Hynard

Student thesis: Doctoral ThesisDoctor of Philosophy


Over the past 20 years, winters in the UK have become milder with the odd harsh winter and with the drive to use low carbon cements, questions have arisen over the performance of these concretes in freeze-thaw environments. This research project investigated the influence of cement type on concretes subjected to freeze-thaw conditions and the correlation between the microstructural properties of concrete and freeze-thaw performance using the CEN/TS 12390-9. The test method based on SS 137244 with a temperature profile of +20±4°C to -20±2°C and the results were compared to a scaling loss of up to 1.0 kg/m2 (deemed Acceptable performance).

Concretes were manufactured with CEM I, CEM II/B-V (fly ash), CEM III/A (GGBS) and CEM II/A-L (limestone) cements, both non-air and air entrained, with different target strengths (20-60 MPa for non-air entrained and 20-50 MPa for air entrained) and a target air content of 4.5%, in accordance with BS 8500. BS EN 197-1 outlines the maximum addition contents that can be used in concretes and BS 8500 describes lower maximum limits for these additions regarding freeze-thaw conditions. CEM II/B-V (45%, 55% and 65%), CEM III/A (65%, 75%, and 85%) and CEM II/A-L (30%, 40% and 50%) were tested with addition contents higher than the allowable limits to determine how these influence the air void characteristics and freeze-thaw resistance. Concretes were analysed to determine the effects of air entrainment on the air void characteristics (air content of hardened concrete, spacing factor, specific surface, void frequency average chord length and microair content) and subsequent freeze-thaw resistance. The study also examined the effect of cement type in concretes with a range of target air contents (7.0%, 9.5% and 12.0%).

Powers (1945) derived the Spacing Factor parameter to determine if a concrete could resist freeze-thaw whereby voids within the concrete were less than 250 µm apart. Development of 3rd generation superplasticizers combined with air entrainers, both air and non-air entrained concretes had a spacing factor value less than 250 µm (and for most concretes <100 µm), however many of the non-air entrained concretes did not achieve an Acceptable scaling rating showing that Spacing Factor cannot be used as an initial assessment of freeze-thaw resistance. Other parameters including the specific surface and void frequency should be considered when determining freeze-thaw resistance as these provide better correlation with performance. For non-air entrained concretes void frequency should be <0.600 mm-1 and where air entrainment is used, >0.600 mm-1 is acceptable.

All air entrained concretes (target air content 4.5%) achieved an Acceptable scaling rating even with the varied cement type and target compressive strength. Non-air entrained concretes with a strength 40 MPa or less did not achieve this rating, and a 60 MPa CEM II/B-V concrete was not able to withstand freeze-thaw damage. Higher air contents were shown to protect the concretes, however there was a plateau point where the air entraining admixtures changes from increasing the air content to altering the workability like a superplasticizer but it did increase the freeze-thaw resistance with a maximum compressive strength loss of up to 25% for some cement types. Increasing the addition content above the maximum decreases the compressive strength. Despite the fact the compressive strength decreased with increased addition contents, the concretes were still able to achieve an Acceptable scaling rating. BS 8500 also states that a lightweight aggregate concrete can perform well in XF4 exposure conditions. This was studied to understand the microstructural properties and whether the aggregate replicates an air entrained concrete in terms of air void size and distribution, and to that end it was observed that lightweight aggregate results were like air entrained concrete.

CEN/TS 12390-9 is based on the Swedish Standard 137244 for freeze-thaw resistance and compared to the UK, these temperatures are rarely seen. Different temperature ranges (+13°C to -13°C and +18°C to -8°C) were considered to reflect the temperatures more seen in the UK highlighting that for a concrete to be warmer (but still below 0°C) for longer, produced more scaling than CEN/TS 12390-9 temperature profile. Moreover, other environmental factors are not considered in the test including the effects of carbonation since concrete is exposed to carbon dioxide all year round then this can potentially influence a concrete’s resistance and found higher scaling loss for carbonated, non-air entrained concretes, in particular CEM II/B-V and CEM II/A-L concretes with a loss of 12.35 kg/m2 and 11.07 kg/m2 respectively. Increased salt concentration from multiple applications was studied to determine how increasing the salt levels (from the standard 3% to 6%, 9% and 12%) would affect the concrete’s durability. Observed was the change in freeze-thaw mechanism between 6% and 9% from surface scaling to internal freeze-thaw cracking.
Date of Award2021
Original languageEnglish
SupervisorMoray Newlands (Supervisor) & Michael McCarthy (Supervisor)


  • Freeze-Thaw
  • Concrete
  • Concrete durability
  • Carbonation
  • Micro-CT
  • Microstructure
  • Spacing Factor
  • Specific Surface
  • Void Frequency
  • Microair
  • Air Content
  • Lightweight Aggregate
  • Salt Concentration
  • CEN/TS 12390-9
  • Air Void Analyser
  • Temperature Profile

Cite this