Power to protective structures

Materials World magazine
1 Dec 2010

Structures that can withstand anything man or nature throws at them are not simply unrealistic fodder for a Hollywood action film. Materials scientists and engineers worldwide are quietly working to make protective construction a reality. Rupal Mehta finds out more.

Having the ability to create reinforced concrete columns that can better resist blasts, or glass windows that do not shatter into flying shards, could have vast implications for civilian and military life, reducing the impact of terrorism, warfare or natural disasters. Yet, it is only now that an international conference dedicated to protective structures has surfaced.

The First International Conference on Protective Structures was held in Manchester, UK, from 29 September to 1 October. It saw nearly 100 papers presented by speakers from around the world, and launched the International Association of Protective Structures (IAPS).

The Association’s first President, Professor Norbert Gebbeken, of the University of Bundeswehr in Munich, Germany, emphasised the need for such an organisation. It will aim to provide a stronger voice for the field when lobbying for research funding, and offer a focal point for knowledge transfer through the Association’s website, its International Journal of Protective Structures and biannual conferences.

Gebbeken explained that good representation is needed at European Commission level, where ‘the majority of [research] funding for safety and defence has gone to the security industry, rather than protective structures, because there was a lobby for the [former]’. Having a voice at international, European and national level, he suggests, can help identify where funds can be spent for the short-, medium- and long-term benefit.

Dr Colin Morison, a delegate at the event and Technical Director of the Security & Explosives Effects Division at TPS Consult, based in Croydon, UK, echoed the need for a firmer platform. Talking to Materials World afterwards, he says, ‘most of the other conferences are military in origin. Yet protective structures are both military and civilian. I can recognise the gap in information that this [event] can help fill’.

Dr Qingming Li, of The University of Manchester, UK, and Chair of the event’s International Scientific Committee, adds, ‘The argument has been that prevention is better than protection, but, as the last resort, to protect people, protection is vital’. The IAPS, he suggests, will not only help to present a unified front but might also plug the gap between researchers and practioners.

Any additional funding that is secured would aid investment in the ‘expensive test facilities’ needed so that ‘a whole string of research can be done’, asserts Morison.

Problem shared is a problem solved?

Keynote speaker Professor Bryn James, of the UK’s Defence Science and Technology Laboratory in Salisbury, illustrated the complex nature of the research that still needs to be undertaken. His presentation was littered with problems rather than solutions. He quipped, ‘Unlike people who have presented their work and solutions, I am just saying what the difficulties are’.

‘These are not simple threats’, he outlined, and to protect against them, you need to understand the physics behind, for example, how bullets, blasts and other weapons penetrate. Protection against each threat is different – what are the impact conditions?

James continued, ‘In low velocity impact, the strength of the material dominates. In high velocity impact, [hydrodynamic behaviour of the material is more relevant]’.

The added problem is that the goal posts keep moving. ‘Weapon designers are ingenious about getting through protective structures,’ highlighted James. ‘They are good at getting all the momentum in a small device.’

Understanding and predicting material failure on a microscopic level is crucial to enhancing its protective properties. James reiterated, ‘Materials do not fail as a continuum’.

On best behaviour

The majority of the papers presented at the conference strived to enhance understanding of material behaviour (see 'Paper trail' below).

One of the key themes was investigation into the disproportionate collapse of buildings. Morison outlines, ‘If you can prevent damage to columns, for example, by reinforcing the ground floor concrete columns so they carry the load even in damaged conditions, disproportionate collapse does not get started. There were a number of papers on this issue’.

But while all the researchers were commended for their work, it became clear from the subsequent question and answers sessions that there are challenges involved in perfecting experimental tests and computational materials models to take the work forward on a practical level.

As a practitioner, Morison is acutely aware of the pitfalls. ‘Some of the work is good and some of it escapes the real point. Some researchers start with simplified loading and then try to do complicated analysis of the behaviour, even if the loading is unrealistic.

‘For example, they try to produce a triangular history of pressure and impulse, uniformly distribute it up the column and then apply it to a complex non-linear LS Dyna numerical model to see how reinforced concrete behaves. With a close-in charge, however, you are going to get non-uniform distribution up and down a column, and a difference in the time of arrival, which is ignored in a simplified loading model. However good the analysis is, you cannot relate the results back to a charge and stand off’.

He adds, ‘With these pressure distributions, what you need to consider is, is this a freestanding column, or a façade of a building, where load from the façade will come into the column’.

Computational conundrum

One of the keynote speakers, John Crawford of Karagozian & Case, in Burbank, USA, cautioned that engineers should not be enamoured by the deceptive ease with which computational models can now be produced.

He explained, ‘Any kind of protective design requires some form of analysis – but when you look at the analysis produced, it is clear we are being pulled into a false sense of security by pretty pictures. It is easy to make models that look good but perform badly, and lead you to make choices that are wrong. Calculations must get their due’.

Engineers working in the sector need to realise that they are ‘going to have a lot of lives in their hands’, before they start ‘touting’ their results, insisted Crawford.

He urged the academics in the audience to instil the need for groundwork among their students, acknowledging that it is ‘hard to slow them down and suggest they look at something in one dimension first’, and then follow this with selection of an appropriate computational model. ‘If you are going to focus on direct shear [in columns], find a model that works well in finding it – the due diligence part of analysis needs more emphasis’.

Gebekken, Crawford and Morison propose that the IAPS could help enshrine a concept of ‘modelling reliability’ by developing a shared and accessible database of materials models.

The ability to achieve this, however, falls back to the original point of funding. ‘It takes a central authority to identify the need and spend the required money,’ insists Morison, ‘otherwise people make academic advances as and when they get grants that only nibble at the edges of the questions. To get the modelling right, you need to have good experimental data to calibrate the model against. Part of the problem is also that the testing equipment is quite esoteric' (see 'Under strain' below on split Hopkinson pressure bar testing).

Spend wisely

It was clear that all delegates recognised that there is much good work ongoing in the field but that now is an opportune time to channel it better.

Mark Stewart, of the Centre for Infrastructure and Reliability at The University of Newcastle, Australia, presented on the need for Governments to prioritise protective measures against terrorism for key
infrastructure by undertaking cost benefit analysis to ascertain how much it reduces the risk, how many lives could potentially be saved, and the probability of that infrastructure being targeted.

Morison, meanwhile, notes the need to balance design with protective material solutions, suggesting the latter might not always be the most economic solution. He points to the new American embassy in Vauxhall, UK, as an example of a building designed with a large compound, with perimeter walls to prevent vehicle bombs gaining access. The stand-off should provide blast protection in conjunction with enhanced glazing.

He continues, ‘Many existing embassies, on the other hand, are on the kerbside with little stand-off. In addition to enhancing glazing and masonry walls, you may look to provide multiple layers of defence from the street by adopting a layout where the front rooms are sacrificial with the important rooms further back. The context in which you apply the materials research is important’. 

Paper trail

The First International Conference on Protective Structures encompassed nearly 100 presentations on a range of materials and concepts. These included –

  • ‘Fibre-reinforced Ultra-high Performance Concrete (UHPC) – a Material with Potential for Protective Structures’. Oliver Millon from the Fraunhofer Institute for High Speed Dynamics, in Freiburg, Germany, outlined research into the material’s behaviour at high strain rates. A comparison between UPHC, high perfomance concrete and conventional concrete has been sought.
  • ‘A Study on Progressive Collapse of Steel-frame Structures under Explosive Loading’. The
  • investigations on four-storey structures were conducted by a team at the Beijing University of Technology, China.
  • ‘Frangible Window Concept for Reduced Casualties in an Explosive Attack’, presented by David Stevens at Protection Engineering Consultants in San Antonio, USA. The concept aims to reduce casualties from flying debris. Instead, a proprietary window has been designed, under the
  • sponsorship of the US Department of Homeland Security, where motion is controlled through a novel retention system.
  • ‘Structural Reliability Analysis of Unreinforced Brick Masonry Walls Subject to Explosive Blast Loading.’ One of a small number of papers on masonry walls, the presentation outlined the use of Monte-Carlo simulation and probabilistic methods that incorporates the uncertainties associated with blast loads and material properties.
  • ‘Explicit Dynamics Simulation of the Natural Fragmentation of Cased Explosives’. The work
  • conducted by ANSYS in Horsham, UK, and the Defence Ordnance Safety Group of the UK Ministry of Defence, aimed to determine the loading from this threat by improving the capability of the ANSYS AUTODYN software.
  • ‘Development of a Lightweight Relocatable Shelter System with High Resilience to Blast, Ballistic and Seismic Threats’. Conceived by a consortium in Canada, the ‘air-beam shelter’ has no hard framing, or panelling and can enclose spans to 40m. The large columns of polyester-fabric tubing are formed into arches and lightly pressurised by air blowers.

Under strain

The complexities of using split Hopkinson pressure bar (SHPB) testing to determine the strain rate effects on materials properties was a recurring theme at the event with several papers presenting the possible misinterpretation of SHPB results and potential corrective methods that could be employed.  In an interview with Materials World after the conference, Dr Qingming Li of The University of Manchester, UK, and Chair of the event’s International Scientific Committee, outlined the problem.

The technique was developed in the late 1940s,’ he explained, ‘and thousands of papers have been published to determine the uniaxial stress-strain relations at high strain rates for various engineering and natural materials. It is a quasi-standard test as no organisation has made this a standard technique’. 

The challenge, he notes, is that the stress state in a sample in a SHPB test may become multiaxial when the strain rate is greater than the critical transition strain rate. But, many researchers still interpret the results as uniaxial stress-strain curve, and implement them into the material model in software.

‘If you treat it as a uniaxial compressive stress-strain curve and put it into the material model, the interpretation is wrong because the measured stress-strain curves are in multiaxial stress state, which leads to lateral confinement.’ asserts Li. ‘[It means that] they may artificially overestimate the strain rate effect and think the material is stronger than it is. That is dangerous for the design of protective structures’.

Li and his colleagues at Manchester have proposed a corrective numerical methodology that follows the SHPB tests to verify the real strain-rate effect of brittle materials such as concrete.


Elastic behaviour

Dynamic retrofitting of existing un-reinforced masonry infill walls was the focus of a presentation by Paolo Casadei, Technical Director of Fidia sri – Technical Global Services in Milan, Italy.

The objective has been to enhance the material’s out-of-plane resistance in case of dynamic loads that may be generated during an earthquake or by a blast. Casadei explained that fibre-reinforced polymer (FRP) laminates are useful for this purpose, but one limitation is ‘lack of ductility at ultimate state’, as well as poor behaviour when impacted by debris. This is based on static, as well as real, blast tests performed on several retrofitted walls.

Fidia’s solution is an elastic system that is said to combine the strength of FRP laminates with the ductility of polyurea resin. It encompasses three layers –

  • A less than one millimetre thick ‘strengthening layer’ of FRP, which enhances flexural out-of-place strength.
  • A ‘dissipative’ section formed by impregnating fibre grids (made of ultra high tensile strength steel or aramid fibres) with polyurea and interspersing with ballistic layers. The thickness depends on the level of the protection required.
  • An ‘elastic layer’ of about 10mm thickness or more. This comprises a polyurea external layer anchored to the main bearing frame to improve the infill masonry walls’ ductility and to provide a ‘catching net’ for any debris that may pass through the other layers. This section slides over the others using a chemical debonding layer. This new protective system has been named Elastic Systems for Dynamic Retrofitting.

The same system is also being explored for fabricating a protective school desk – the Safety Desk – entirely from such materials. The idea came to the firm following an earthquake in 2002 that killed schoolchildren at the School of San Giuliano di Puglia, 140 miles southeast of Rome, Italy.

Casadei explains, ‘Usually in an earthquake, teachers tell schoolchildren to go under their school desks to protect themselves from the ceiling collapsing, but the conventional school (wooden surface and steel frame) desk is not strong enough’. He states that although the new desk is likely to be about 15kg (a few kilos heavier), it ‘is still below the 25kg limit that workers can carry’, and is said to be 400% stronger based on drop tests of loads from three, metre heights. Research is ongoing.


Further information


See also from this issue: On the counter-offensive - designing for counter-terrorism


Documents for download: 
Image icon p6_mw0109_oil.jpg174.65 KB