Changes to the beam geometry, slab system, connections, loading or required fire resistance can all take a design beyond the precise scope of the available test evidence.
This relationship between testing, engineering analysis and project-specific design was at the centre of Designing for Fire, Peikko UK’s recent technical event at The Building Centre in London.
The hybrid seminar and webinar brought together Mikko Salminen, Ph.D., Structural Fire Engineering Lead for Europe at Jensen Hughes, and Vesa Miettinen, Senior Manager, R&D for DELTABEAM® at Peikko Group Corporation.
Through two complementary presentations, they explored how structural behaviour in fire can be understood using evidence from physical testing, validated analytical models and the wider behaviour of the structural system.
Standard fire tests provide a starting point
Fire testing remains fundamental to structural fire safety. It allows engineers and manufacturers to observe temperatures, deformation, load-bearing capacity and failure mechanisms under controlled conditions.
However, Mikko explained that standard fire curves provide a consistent basis for classification rather than recreating the development of every possible real fire. A 60-minute classification describes performance in a standard test - it doesn’t automatically mean that a structure will resist every real fire for precisely 60 minutes.
Physical tests also have practical limits. They normally cover a specific specimen, loading arrangement and set of boundary conditions. It would be prohibitively expensive and time-consuming to test every variation subsequently proposed on a project, as Mikko explained:
“We cannot test every possible arrangement. The important thing is to complete the required assessment at a consistent technical level and begin that process as early as possible.”
Engineering assessment becomes particularly important when the proposed structure moves beyond the available test evidence.
Structural fire engineering considers the wider system
Prescriptive design methods commonly assess individual elements using standard fire exposure and established calculation procedures. This approach is appropriate for many buildings, but more detailed analysis becomes valuable when a structure is unusual, contains untested details or has the potential to behave differently as a complete system.
Performance-based structural fire engineering can combine realistic fire scenarios with thermal and structural analysis. Depending on the objective, this might range from a focused assessment of a connection or detail to a global model of the structural frame.
The aim is not to use the most advanced model available simply because it exists. The analysis should be proportionate to the design question and supported by appropriate validation.
This wider perspective can reveal behaviour that an isolated element check may not capture. Composite floors, for example, can undergo considerable deformation while continuing to carry load through alternative mechanisms. Steel positioned well away from a likely fire may experience lower temperatures than a uniform prescriptive assumption would suggest.
These effects can sometimes support a more targeted fire-protection strategy, but only where the structural behaviour has been properly assessed and the assumptions are understood by the wider design team.
Fire behaviour can continue into the cooling phase
One of Mikko’s examples involved a cross-laminated timber slab exposed to a natural fire.
The test continued considerably longer than the initial analysis had predicted, but failure eventually occurred during the cooling phase. Although the fire exposure was reducing, heat was still moving through the timber section and affecting its remaining capacity.
The example demonstrated why structural fire analysis cannot always stop at the point of peak temperature. Cooling can form a significant part of the fire scenario, particularly where materials retain heat or experience continuing thermal penetration.
Test data from the experiment was subsequently used to calibrate thermal and structural models, allowing the behaviour to be studied in greater detail.
This process illustrates the relationship between testing and analysis. The test reveals how the structure behaves, while the model helps engineers understand why it behaved that way and assess related arrangements.
DELTABEAM® evidence has developed over decades
Vesa’s presentation showed how the same relationship has shaped the development of DELTABEAM®.
The first DELTABEAM® fire tests took place in 1990. Since then, Peikko’s research programme has expanded to cover different beam configurations, slab systems, fire ratings, connections and load-transfer mechanisms.
Testing has included individual beams, full-scale beam-and-slab assemblies and arrangements using hollowcore slabs, solid concrete slabs, steel decking and timber floor systems.
“You cannot test every beam size and configuration. The purpose of a test is to establish principles that can be applied across the related cases and beam sizes that will be encountered on real projects” as Vesa described.
The results provide temperature profiles and structural capacity data that can be used to validate numerical models. Once validated, those models allow engineers to assess related configurations that cannot all be reproduced through full-scale physical testing. The evidence also feeds back into product development, design methods and PeikkoDesigner®, helping ensure that project calculations reflect lessons learned through testing.
In most cases, the physical tests have demonstrated performance beyond the conservative analytical prediction.
“In general, DELTABEAM® has performed better in testing than we expected from the analysis. In North America, for example, we did not expect one test to reach four hours, but the beam continued to carry the load”, explained Vesa.
That margin is reassuring, but the purpose of the test programme is not simply to achieve the highest possible rating. It is to understand how the system carries load, where its capacity comes from and how reliably that behaviour can be applied in design.
Applying international evidence to a UK project
A current London office project provided a practical example of how evidence can be transferred between tested and proposed arrangements.
The project required 90 minutes’ fire resistance for a DELTABEAM® and metal-decking configuration that had not been tested to the relevant British standard. Test evidence was available from North America, but differences between the tested arrangement and the UK proposal still needed to be addressed.
Jensen Hughes first modelled the tested configuration and checked that the analysis reproduced its thermal behaviour conservatively. The validated method was then applied to the proposed structure, accounting for differences between the test standards and project conditions.
The results supported the required fire resistance and were consistent with the temperatures calculated by PeikkoDesigner®. Despite that analytical evidence, the project team and approving stakeholders wanted further physical confirmation. Peikko and Jensen Hughes are therefore progressing additional testing to strengthen the UK evidence base for this type of arrangement.
Rather than presenting testing and modelling as alternatives, the project shows how they can support each other. Existing tests inform the analysis, the analysis addresses project-specific differences, and further testing can provide additional evidence where it is needed for approval and future applications.
Early conversations create better options for optimisation
The practical message running through both presentations was the importance of addressing structural fire engineering early.
Late analysis may still identify a workable solution, but the available options are narrower once the structural design, procurement route and fire-protection strategy have been fixed.
Early involvement allows the team to establish the objective, identify gaps in the available test evidence and agree an appropriate assessment route. It also creates time for discussions with approving authorities and other stakeholders.
“If a steel structure is positioned well away from the potential fire and the requirement is one or two hours, it is worth involving a structural fire engineer early. There may be potential to optimise the protection, but that conversation needs to begin while the project is still at an early stage”, said Mikko.
For projects involving exposed steelwork, large composite structures, timber systems or unusual connections, this can make a substantial difference. It may prevent an untested detail from becoming a late-stage problem and identify where protection can be concentrated rather than applied uniformly.
The right solution will still depend on the building, its fire strategy and the evidence available. However, testing, validated analysis and early collaboration give project teams a stronger basis for making those decisions.
The event was hosted and moderated by Darren Waller, Managing Director of Peikko UK.
We would like to thank Mikko Salminen and Vesa Miettinen for sharing their expertise, and everyone who attended in London or joined us online.
