toheroa jim: blisters on asphalt pavements

A previous blog made reference to blisters forming within ice that had formed on the reservoir in Central park, New York, during the cold snap of February, 2010. What immediately intrigued me was that these ice blisters were very close in form, and also very possibly in their thermo-mechanical origins, to some earlier blisters in asphalt pavements that had caught my attention.

In the course of an unusually hot summer some years ago a number of curious bulges formed within the thin asphalt layer that had only a short time before been laid over the footpaths on Gower Street, just a few metres away from my office at UCL. This practice of applying a thin coating of asphalt appears to have been initiated by the London Borough of Camden (LBC) to overcome the longstanding problems associated with the previously used blocks of paving stones. Hitherto, differential movement between the paving stones on the footpaths tended to develop serious discontinuities between adjacent pavement blocks upon which the unwary pedestrian was liable to trip. As a consequence, local authorities (and certainly not just LBC) whose responsibilities include the maintenance of the footpaths were being faced with expensive compensation claims. Replacing these paving stones with smooth asphalt layers seemed like a good idea. It would remove these discontinuities and consequently reduce the risk of accidents. Well that at least was the idea until the appearance of the upward blisters.

The photograph above show a number of these blisters formed on what had a few weeks previously been an ostensibly flat pavement. Most can be seen to have an effectivelycircular, axisymmetric, geometric form. In some cases these circular blisters have started to overlap, creating little clusters of blisters. In one or two places there can be seen to be nasty scars which represent the remains of LBC's attempts to remove the previous season's crop of blisters. Convinced that these blisters were the result of thermally induced upheaval buckling I included an analysis of their possible development in a short article (written in 1999) discussing the mechanics that would underpin such thermally induced buckling modes in continuous pavements. This was not difficult because it was merely a 2-dimensional version of some earlier work I had been doing to predict the possible occurrence of upheaval buckling of subsea pipelines when hot oil or gas is pumped through them. Incidentally this major problem for oil and gas companies involved with offshore operations is very closely related to the lateral buckling that can occur in railway track when exposed to high summer temperatures. But I will return to some of these closely related problems in later blogs. For the moment let me concentrate on the asphalt pavement blisters.

A few months after preparing the little piece analysing the pavement blisters as a form of thermal upheaval buckling, I decided to include it in a talk I was giving at the University of Hong Kong. Unexpectedly, since this is not a topic for which one would expect a great deal of excitement, this aspect of the talk attracted quite a lot of fairly animated discussion. What I had not realised is that a week or so before the talk the asphalt “wear layer” covering the steel plate for the newly completed Chiang Mai suspension bridge had developed similar but apparently somewhat larger blisters. Again, not a subject that one would have thought to be worthy of a great deal of animation, except for the fact that Queen Elizabeth II was due to open the bridge a week or so later. It was though in the course of this discussion I came to realise that what I thought would have been a fairly widely recognised explanation for the origin of pavement blisters was not the one to which pavements engineers had come to subscribe.

Conventional wisdom seems to have formed the view that these blisters within thin asphalt layers are the result of gas pressure building up beneath the aphalt sheet. Not altogether a mad idea given that many upward bulges in asphalt pavements are so often the relief of the pressures caused by the growth of roots of trees or other plants. There were certainly no tree roots on the Chiang Mai bridge deck and I am confident there were none beneath the pavements on Gower Street. But the gas pressure variant of this model seems plausible. There are at least two schools of thought as to the origins of the gas pressure. When the black asphalt surface is subject to radiant energy from the sun it and the underlying soil can experience very high temperatures. In these circumstances some of the moisture in the underlying soil might be expected to absorb sufficient energy to allow it to evaporate to form steam. If totally restrained this phase change would be accompanied by the development of high gas pressures. So what better way to relieve these high pressures than to induce an upward bulging in the form of blisters in the asphalt layer? A variant of this gas pressure explanation of the blisters is that the gas release is the result of the evaporation of certain volatile materials within the asphalt mix itself. This source of gas pressure would certainly be much more credible in situations such as the Chiang Mai bridge deck where it is unlikely that much moisture would exist between the steel top plate of the box girder and the asphalt wear layer.

However, there are a few problems with the gas theory expalanations. Blisters will form in those areas where the bond between the asphalt and the underlying material is relatively weak. Areas of low bond strength will most likely have fairly random shapes. That being the case the plan shapes of the upward deformation of the asphalt if formed through gas pressure would reflect the probable irregularity of the weakened areas of bond. Laws of probability would therefore suggest a circular planform to be the exception rather than the rule. And yet the plan shapes of the blisters as shown in the above typical photograph, and incidentally many more taken from asphalt pavements around the world, have a very robust tendancy to form with almost perfectly circular planforms. A related problem is that the actual extent of the weakened bond areas would be most unlikely to be the same at different locations over the asphalt surface. In these circumstances a gas relief explanation would be expected to produce plan areas of the blisters that reflect this irregularity in size. Again, this does not seem to be the case. Blisters of a given amplitude of uplift at different locations over the asphalt surface have a remarkably consistent lateral extent. Furthermore, in a gas pressure relief model the lateral dimensions of the blisters would be unlikely to change very much as the amplitude of the upward bulge increases. Again, this does not conform with observations that show the diameters of the circular bulges increasing as their heights increase.

But if all this is too abstract to convince you that the origins of the blisters might be something other than the relief of gas pressure there is a much more compelling piece of evidence that might help. Blisters tend to grow incrementally over hot periods measured in days to weeks. Growth rates depend upon both the temperature ranges and the numbers of cycles of heating and cooling experienced. By selecting two embryonic blisters of around the same degree of growth a tiny hole was drilled into one but not the other. The hole penetrated into the void beneath the asphalt blister but was small enough to be barely visible. Over the next week or so the blister with the hole was found to grow at the same rate as the one without a hole!