Human induced loading on staircases

Over the last decade it has become increasingly popular to provide large public areas with gracefully designed 'flexible' staircases. One inherent characteristic of this type of construction is a low stiffness to mass ratio and hence a low natural frequency when compared to more traditional designs. A number of staircases have been found to be dynamically responsive to pedestrian traffic resulting in costly repairs. The objective of this thesis was to investigate the differences between human induced loading on floors with that on stairs. Experimental work carried out on a purpose built staircase showed that forces up to 3 times the static body weight were generated during fast descents while forces up to 2.5 times the static body weight were generated during fast ascents. The work also showed that first harmonic values generated while ascending were slightly higher than for descending whilst second harmonic values were up to 3 times greater for fast descents than for fast ascents. When compared with floor testing, stair testing produced first harmonic values nearly 2.5 times greater with second, third and fourth harmonic values nearly 3 times greater. The harmonic results for the flat testing were also incorporated into a new mathematical expression to predict peak accelerations on simply supported floors and footbridges. The experimental results were duplicated analytically by developing a computer program to calculate the vertical ground reaction forces from body segment positional data. Following a Newtonian approach, the predicted first harmonic values were 20% to 30% lower than actual while the second harmonic values were approximately the same. Monte Carlo simulation techniques were also used to model the effects of group loading on stairs. The simulations predicted enhancement factors (a multiplier on single subject loading) of 3 to 6 for smaller groups( 25 people). If the experimental/analytical results are combined with the group loading predictions, the harmonic values for groups ascending or descending flexible staircase could be substantially increased. These results demonstrate that loading data from floors is highly inappropriate for staircase design