Research into the dynamics of structures has advanced the understanding and numerical modelling of the vibration of hospital floors, which is permitting the use of more economic forms of construction.

Traditional thinking among structural engineers has it that hospital buildings, particularly theatres, have to be heavy concrete structures to ensure they meet stringent vibration limits. Due to the results of extensive research into the dynamic behaviour of hospital buildings however, that line of thought has now been superseded.

Advanced numerical modelling of structural vibration has been developed, backed up by physical testing, which has given a substantial boost to the economics of hospital design. Relatively lightweight steel frames and slender post-tensioned concrete and steel-concrete composite floors are being used, now that proven analytical data can be generated to support them.

"Hospitals are complex structures and have very specific requirements with regard to vibration. They are also all different and a long way from relatively simple office type buildings," says Carl Brookes of engineering consultant Gifford. Brookes heads up Gifford's Engineering Analysis Team, which is working with hospital designers to check that their designs will perform as required in terms of vibration.

"There has previously been a lack of understanding of how buildings vibrate in response to dynamic loading, which in hospitals is essentially people walking in rooms and corridors," he says. "This has led to somewhat conservative designs and hospitals historically being built with thick and potentially uneconomic floor slabs."

Gifford set up its own working group to study the structural vibration of buildings in 2002. At around the same time, the Steel Construction Institute (SCI) was carrying out vibration tests at a Gifford designed hospital in Yorkshire as part of a research project aimed at updating long established design guides.

"Fire damage and a consequent process of retrofitting steel and concrete composite floors at this particular hospital allowed the SCI the rare opportunity of measuring vibrations set up by walking tests in an empty hospital with various levels of fit-out," Brookes says.

"The opportunity to measure vibration on bare floors, which are the least dampened and therefore worst-case scenario, as well as fitted-out floors, ensured SCI got a range of results that reflected the extremes of future possible building usage."

The result of the SCI's work is a new hospital design guide that is helping the steel sector make in-roads into a market traditionally dominated by concrete. The SCI also made its test results available for Gifford to use for verifying its numerical modelling software, the development of which is key to the consultant's work with hospital
designers.

Brookes says: "Our work started with looking at composite floors and has now also involved modelling vibration effects in complex structural frame arrangements and post-tensioned concrete floors. Engineers are pushing forward hospital designs, with floors that are stronger for a given depth, so spans are getting bigger and floor depths
are being reduced. The economics are better, providing the designs work well dynamically."

It seems the move to privately financed developments has given a boost to more advanced hospital designs. PFI consortia and health care trusts, with a keen eye on costs and risk, are looking for more efficient forms of construction and assurance that innovative designs will meet performance regulations.

According to Brookes, Gifford's clients are also keen to know how their buildings will perform if walls and partitions are moved, as a way of assessing how much flexibility designs offer for future changes of use.

"We are now applying a lot more rigour to the modelling and analysis and using a more holistic approach," he says. "The whole floor is analysed and instead of using specific walking patterns, a system that is equivalent to walking at every part of the building in turn has been developed so that all possible walking scenarios are considered."

Gifford's software involves automated processing of results from finite element models of the floors including the necessary frequency weighting for use with human perception calculations to show what Brookes describes as, "a contour plot of worst case scenario vibration responses". The end product is a coloured plan of where the floor is within or outside HTM2045 requirements.

"The response factor limits given in HTM2045 of 1.4 for wards and 1.0 for surgical theatres compare with a typical office building response limit of 4.0," Brookes says. "Vibration is among critical criteria in hospital design, but it is traditionally the area least understood. The latest steel and concrete composite and post-tensioned concrete floors are slim, more highly stressed and with fewer structural members. They are very economic, but they also tend to be a bit more lively."

"We are effectively providing a serviceability check and showing in a user-friendly way, a picture of where the building will and will not perform as required. We can suggest solutions if the design is outside response limits, although we are often finding that the floors are meeting HTM2045 requirements and could be designed even thinner.

"Historically, engineers have been applying very simple calculations, such as rules based on natural frequency alone, whereas it is the amplitude of acceleration that is key."

According to Brookes and for those of a technical inclination, the amplitude and frequency of vertical acceleration of a structure's vibration are important because this is what is felt the most due to the body's natural frequency. When standing they are felt acutely in the frequency range of 4Hz to 8Hz.

To predict a structure's vibration behaviour, or 'response', RMS (Root Mean Square) acceleration values are calculated where vibration is considered to be continuous. These figures, which are effectively averages through time, are used to assess the magnitude of vibration by applying weighting curves. Weighting normalises the vibration across a range of frequencies so, for the evaluation of human exposure, their perception is the same at different frequencies.

For example, people standing on floors feel frequencies below 4Hz or above 8Hz less and so their influence in weighted RMS values is reduced. Different weighting curves are used according to different body orientations and postures to complete the process for human evaluation. Similar procedures are required for sensitive equipment such as MRI machines and further procedures are used if vibration is not continuous.

The outcome is a set of vibration responses. They will only meet the current response factor limits if the ratio of calculated weighted RMS acceleration is less than or equal to the requirements given in HTM2045.

"Structural dynamics at this level requires complex data manipulation and the use of signal processing algorithms generally reserved for acoustic and electrical engineers." says Brookes. "Structural engineers are usually concerned with deflections under relatively large and constant loadings, or in design for seismic conditions, much larger amplitude of vibration.”

According to Brookes, Gifford has developed an approach to align with the requirements of HTM2045, but there are a number of areas where better specifications are required, in particular with regard to loading. Better information on the requirements of sensitive equipment is also required.

“It is important that HTM2045 reflects current and as far as possible future vibration requirements to avoid vibration becoming intolerable." adds Brookes. “Given good information on building usage and vibration requirements, we now have the tools to ensure our designs perform efficiently.”

Back

Valuable Insight into Floor Vibration
Gifford
Published in Health Estate Journal, April 2005