Development of a floor vibration design method for Eurocode 5

More comparison calculations and measurements areneeded to determine appropriate vibration criteria.This is underway in CEN TC250/SC5/WG3/subgroupfloor vibrations.It has been decided to keep the stiffness criterion.For very light floors less than 10 x the walker massthe loading models are no longer valid so it makessense to keep the point load deflection limit in placefor these floors. The stiffness criterion allows for anapproximate comparison of the proposed performancelevels with existing national requirements. Howeverfurther work is needed to ensure the stiffness criterionand the response factors correspond.The dynamic response of floors is highly sensitiveto the modal mass of the floor. A higher modal massis very beneficial being directly proportional to theinverse of the response as long as its effect on thefrequency is ignored. Therefore, if a realistic levelof live load is included in the mass, it will have abeneficial effect on the calculated dynamic response.In the calculations presented in this paper live loadshave not been included, but this is under discussion.It can also be seen that wide rectangular floors havea lower response than square floors of the same spanand floor build up due to their increased modal mass.Discussion is ongoing of what walking speed (frequency)is appropriate. It is clearly recognised that smallerfloor areas result in lower walking frequencies as isconsidered in the Finnish national annex. The walkingfrequency has a direct impact on the dynamic loadapplied. A simple proposal being considered is that a1,5 Hz walking frequency is used for residential spacesand a 2,0 Hz or greater frequency is used for officesand other uses. This may be oversimplified and oneoption would be to link walking frequency to the floordimensions and walking patterns possible on longerstretches. For very long corridors etc walking speedshigher than 2Hz are possible.Another point that is under discussion is that in orderfor floor vibrations to be perceived there needs tobe both a walker producing the footfall and anotherperson sensing the vibration, usually sitting on a chair.These people cannot be at the same point at midspan.Therefore a reduction factor of about 0,7-0,9 has beendiscussed to account for this. In reality this factor isthe mode shape value at the point of excitation bythe walker multiplied by the mode shape value at thepoint of vibration perception by the person sitting onthe chair.Work is currently underway to evaluate suitablemultispan factors to account for the effect of floorbeing multi spanning when evaluating response.Multiple spans mean some increase in the modal massand where spans are not of equal length an increasein the frequency of the floor. The two sets of factorsbeing considered are those in the Austrian nationalannex based on beam theory assuming response ina single mode and those in Wilford et al based onplate theory whilst also accounting for multimodalresponse.Overall it is believed that the method discussed inthis paper offers a simple reliable way to check thevibration of rectangular timber floors on wall supportsthat can be easily built into software and design tools.Work is ongoing to develop a similar method for floorson flexible beam supports.Floor, vibration, design, vibration criteria, acceleration, velocity, frequency, serviceability