Fire resistance - is it a problem for UHPC?
Fire resistance is one of the attributes that distinguishes concrete compared to other materials such as steel or aluminium – but concerns are often raised regarding the fire resistance of UHPC. The short and simple answer to those concerns is that it is quite possible to design a UHPC with a fire resistance that is at least equal to that of conventional concrete – and for CRC i2® it only requires sufficient maturity (to reduce the moisture content) and in a few cases some additional reinforcement.
Fire resistance and spalling – 2 separate problems
The major concern for UHPC under fire exposure is spalling – and actually this can be a problem even for concretes in the range of 70-80 MPa. I will address this below, and take the easiest part first, because even if spalling is not a problem, there are a few other areas where the behaviour of UHPC differs from conventional concrete under fire exposure.
As UHPC is much denser than conventional concrete (lower water/powder ratio and less entrained air) the thermal conductivity is higher than for conventional concrete, meaning that a high temperature in the concrete will be reached sooner in UHPC.
The specific heat capacity is typically lower for UHPC (again a denser material and with lower moisture content). This in itself will result in a lower thermal capacity and as UHPC is often used in very slender elements this lowers the thermal capacity of the element even further, as there is less mass to absorb the heat.
Low cover to the reinforcement
As CRC always uses conventional reinforcement together with steel fibre reinforcement the cover to the reinforcement should be considered. Usually this cover is as low as 15 mm, which means that if the tensile side of the element is exposed to the fire, the reinforcement could reach high temperatures relatively quickly compared to conventional concrete, which would have a much larger cover. This may not be a problem for other types of UHPC, which often uses a combination of steel fibres and prestressing.
Properties at high temperature
It has been shown (eg. with tests at Imperial College assisted by experts from University of Milano) that compressive strength and tensile strength is maintained much better in UHPC at high temperature (up to 700 degrees celsius) than it is in conventional concrete. This is true when measured hot, but even more so when residual strength is measured. This is because of the negative influence of calcium hydroxide, where the amount of calcium hydroxide in UHPC is much lower than it is in conventional concrete.
To sum this up – there is definitely a challenge for UHPC in this area! The way we have addressed this for CRC is to carry out a number of tests (on columns, wall elements and wedge-shaped beams) where the temperature increase at different depths in the concrete (10 to 50 mm) has been measured during fire exposure and this input is then used in fire calculations. In some cases it may be necessary to add extra reinforcing bars to ensure an adequate fire resistance. To test the fire calculations different elements have then been exposed to a full-scale fire test – under load. An example of a test set-up is shown in the picture below. This was a test carried out at the University of Tampere in Finland, where the cantilevered balcony sustained the standard fire for 2 hours before the deformations became too large for the test set-up.
A cantilevered CRC balcony slab being installed in the furnace prior to testing.
There are different kinds of spalling under fire exposure, but the explosive spalling typically associated with UHPC – and experienced in a number of fires such as those that occurred for the Great Belt Link Tunnel in Denmark and the Channel Tunnel between England and France – is caused by water vapour pressure.
When the concrete is exposed to high temperatures the water inside the concrete turns to steam and this steam will try to escape. This is not a problem in a porous concrete, but if the porosity is low, a relatively high pressure will be built up as the steam tries to escape. If the pressure exceeds the tensile strength of the concrete, the concrete will crack in order to release the steam. If the tensile strength of the concrete is high – and the porosity is low – the build-up of pressure can be so large, that when the tensile capacity is exceeded and the concrete spalls this will occur in an explosive manner. There are a number of parametres that indicate whether explosive spalling is likely to occur and some of the factors that increase the risk of explosive spalling are:
One of the first tests to check for risk of explosive spalling was actually carried out by the inventor of CRC – Hans Henrik Bache – in his own kitchen. He put a piece of young CRC – a few days old - on a pan and heated it on his stove. Suddenly the piece of CRC exploded and small pieces were scattered around his kitchen – which gave Bache a very solid idea of the risk.
Since then an extensive series of tests on CRC followed - damaging a few heating elements in furnaces in an attempt to establish exactly where the limits were. The tests demonstrated that if moisture content in CRC is sufficiently low (around 2%) there is no risk of spalling. This can be achieved through drying or simply by having a sufficient maturity, as the cement content is high enough (a water/binder ratio around 0.16) to ensure self-desiccation, and the content of calciumhydroxide is very low. For other types of UHPC – with a lower porosity than CRC – it may be necessary to add polypropylene fibres, as this decreases the risk of explosive spalling.
The full-scale fire tests have also provided information regarding the risk of spalling as has a couple of actual fires where CRC has been involved – all of them during the construction stage. Tommy will briefly present these actual fires in a later post as I am afraid I have already been less than brief in this post. I will just include the picture below as an appetizer to Tommy´s post. In my defense this is a very complicated problem and although I have simplified matters as much as possible I have not been able to do it in a few, short sentences.
Let me know if you have any comments on the matter of fire resistance of UHPC – or if you have some practical experience in this field (perhaps from an actual fire). Every type of UHPC will be different and although large research projects have been carried out (we have been involved in a few) it is still not possible to predict exactly what the risk of explosive spalling will be for a particular UHPC without carrying out tests.
Fire during the construction phase at Heilig Harn in Holland.