Traditionally concrete has often been “defined” by it’s compressive strength – a 40 MPa concrete or a 90 MPa concrete are expected to have very different properties. This has changed with the advent of Ultra High Performance Concrete (UHPC) where other properties or attributes can be just as important – or more important – than compressive strength. And while we’re on the subject of compressive strength…
Characteristic compressive strength
CRC – the type of concrete used by Hi-Con – typically has a compressive strength measured on 100x100 mm cubes of 150 MPa, but as our design strength relates to compressive strength measured on 150x300 mm cylinders, we use a characteristic strength of 110 MPa. Using the same type of cementitious matrix we could also achieve 300 MPa, but that would necessitate the use of other types of aggregates and heat curing. For most of our applications we would not get much benefit of this increased compressive strength – and it would make our concrete much more expensive and thus it would be more difficult for us to compete with other products.
I have often been asked if we can supply a 200 MPa concrete for a particular project, and then – when we go into more detail – it usually turns out that there is actually no need for the high strength. It is just assumed that this high strength means that durability is better, bending strength is higher etc. This is not necessarily the case – and often there are cheaper and easier ways to improve these properties than through higher strength. For most of our products – even though we use the “low” characteristic strength of 110 MPa – strength is not the determining factor for the design. The design is mostly governed by stiffness, reinforcement and cracking stress in tension as Tommy has addressed in his blog (link).
This has been our experience – that we don’t get a lot of benefit with a higher compressive strength in our typical products such as beams, balconies and staircases – but we welcome any comments on this. For other types of UHPC a higher strength may make perfect sense, e.g. combined with prestressing.
Other mechanical properties
As I have mentioned elsewhere (link) we basically use just one formulation for our CRC for reasons of documentation. We then change the type and content of steel fibres depending on how heavily loaded our elements will be to ensure, that we don’t get macro cracks. For special applications we may use other fibres, but the steel fibres are typically best suited for our products as they have good bond, strength and stiffness. In the figure below is shown an example of the results of a standard test according to EN 14651 (a bending test on a notched beam). We have tested one mix with short steel fibres and another mix with a type of fibres that is not really well suited for CRC. We are well aware that other UHPC’s use longer fibres, but in our case – as we always use rebars - we have limited the fibre length to correspond with the cover to the reinforcement, which is typically 15 mm. The only value from this that we actually use in our design calculation is a characteristic cracking stress of 5 MPa, but generally we prefer to see strain-hardening and high bending stresses in curves like the one below. Including the tensile capacity of the concrete in the design would not really give much added benefit compared to the contribution of the rebars.
Young’s modulus is somewhat related to the compressive strength of the concrete, but very dependent on the type of aggregate we have and we typically achieve a value of around 45 GPa. This is not very much higher than what you achieve with conventional concrete, and as most of our structural elements are relatively slender, stiffness is often the governing design parameter. As for the compressive strength, we could use other aggregates to increase the stiffness to perhaps 75 or 80 GPa, but the cost of this would most often be prohibitive.
With the slender elements that we use it wouldn’t be very effective if we had to use cover layers to the reinforcement of 30 or 40 mm. We take advantage of the low porosity and good durability of CRC and use a cover of 15 mm to the rebars. This cover is actually to ensure a good fire resistance as we could use an even lower cover of 8-10 mm if just durability is considered.
Durability testing has included a number of tests such as mercury porosimetry, nitrogen adsorption and micro calorimetry, but the main tests that are carried out with regard to documentation are for carbonation, for freeze-thaw and a special test for chloride intrusion, where reinforced beams are actually loaded during exposure to salt water. This is to check if the loading level has an influence on chloride intrusion as micro cracking increases. The short version of the results is, that for carbonation and chloride intrusion, the rate is so low that it is difficult to actually determine a maximum value for design life – but it is higher than 200 years. For those with a particular interest in these kinds of things, the effective chloride transport coefficient is lower than 5×10-14 m2/s – and other tests (where salt water was used as mixing water) has shown that even with chlorides present next to the rebars, corrosion will not take place as there is no transport of water and oxygen. For freeze-thaw testing, there is no measurable damage.
Optimizing the mix composition
A number of other properties are very important for CRC, but I apologize for already being too long-winded so these will be addressed in other posts – and they include workability, ductility and bond. I will just conclude by mentioning that we are very careful about optimizing on our CRC composition. One reason is that we want to be able to use the large amount of documentation that we have accumulated over the years and another that we have an advantage in knowing our concrete so well. If we change some parametres it may influence other parametres in a manner that was not intended.
For instance, we could easily reduce the porosity of CRC, but this may lead to problems with regard to fire resistance. With our current mix we have no problems with explosive spalling and we can achieve the necessary fire resistance of our structural elements, but if we reduce porosity we may need to add polypropylene fibres to achieve this.
Another example relates to workability and the “sticky” nature of our mix which can make it difficult to achieve a nice finish to an “open” face in the mould. If we produce a mix with reduced viscosity we could have problems with achieving the fibre distribution that we aim for.
These examples just to explain that laziness is not the only reason we have changed so little about our mix over the years. We actually test the limits all the time – and the results could be used in other types of products – but for the moment we are quite satisfied with what we have – and we are usually able to work around the limitations.
I apologize again for the length of the post – but this is a subject of particular interest for me. If you made it to the end and you have questions or comments, please let me know.