Progressive crowd collapse

I've developed modelling tools for the analysis of progressive crowd collapse due to persistent standing in seated areas at football matches (Premier League Project). The modelling suite allows me to construct 3 dimensional areas (stands) and experiment with varying the seat depth, height, spacing, demographics and slopes. Basically, this is crash test dummies for crowds. The prototype tool is shown below.

In the modelling tool we can increase the rake angle (slope) and add/remove barriers (front of crowd) to test the effects of a progressive crowd collapse. Although a relatively crude modelling tool this help assess the nature and dynamics of risk of progressive crowd collapse due to persistent standing in seated areas.

The above models used articulate "ten-pin" model for the individuals and had only two basic configurations (chess board and offset). 

New physics engine

The physics engine and graphics interface, front end and analysis system were reengineered to enable a free-standing crowd and this allows a new type of test for pressure during a crowd collapse. This new physics engine has excitation and energy outputs to enable the study of the crowd pressure during major incidents. The physics engine was also redeveloped to allow for more sophisticated testing and data analysis. Below is a movie files from the pressure analysis tool. In this system we can configure agents, barriers, and extract individual pressure characteristics from the model. This allows me to test novel barrier configurations. 

Below - design screen (green triangles are the virtual camera's - we can observe the crowd from multiple angles and experiment with progressive crowd collapse in both static and moving crowds.

Below - the output from the enhanced physics crowd modelling tool.

Using this tool we can test barrier design, layout/configuration and the ingress packing density. How the crowds fill the front of stage (over time and space) and behaves can be tested for new and novel barrier design/configurations.

Below - video image of the types of crowd packing and excitation/lunge we can test in virtual environments.

Above and below - a concert excitation and pressure analysis. We have been correlating beat frequencies, barrier pressures and pressure differentials within the crowd. Click here for more information.

One of the key elements of crowd modelling is to understand the capacity of the space, how quickly it will fill and what time it will take to reach critical density. This does not need complex computer simulations as simple first pass capacity analysis gives us clear indications whether a system will work (or needs a radical design change).

Some citations on crowd forces

1. Tolerable force against 100m flat bar 623N (Men) up to 800N when they could push back

This is extracted from : "Crowd Pressure Monitoring" by IHG Hopkins, SJ Poutney, P Hayes and MA Sheppard published in "Engineering for Crowd Safety" RA Smith and JF Dickie (Editors) in 1993 by Elsevier.

2. Dangerous Crowd Pressure

“When pushed against a 100mm wide flat bar, the tolerable force was typically 140lbf (623N) for men. When allowed to push back against the bar, to reduce loading on the rib cage, this force increased to typically 180lbf (800N). The tolerable force or women was reported to be significantly less. ....

.... the method was essentially subjective in that it established the discomfort or anxiety threshold of the volunteer.”

2. Crush Asphyxia 6227N for 15 seconds

3. Crush Asphyxia 1112N - if sustained for 4-6 minutes

Also from 

"Crowd Pressure Monitoring" by IHG Hopkins, SJ Poutney, P Hayes and MA Sheppard published in "Engineering for Crowd Safety" RA Smith and JF Dickie (Editors) in 1993 by Elsevier.

2. Dangerous Crowd Pressure

“Firstly, death was estimated to have occurred 15 seconds after a load of 1440 lbf (6227N) was applied. In the second case it was estimated that death occurred after 4 - 6 minutes of applying a load of 250lbf (1112N).

4. 3 people pushing can exert a force of 1370N

5. 5 people pushing can exert a 3430N force

are from “The Causes and Prevention of Crowd Disasters” by John J. Fruin, Ph.D., P.E. United States of America. (Originally presented at the First International Conference on Engineering for Crowd Safety, London, England, March 1993. Elsevier Science Publishers B.B. © 1993)


“Crowd forces can reach levels that almost impossible to resist or control. Virtually all crowd deaths are due to compressive asphyxia and not the "trampling" reported by the news media. Evidence of bent steel railings after several fatal crowd incidents show that forces of more than 4500 N (1,000 lbs.) occurred. Forces are due to pushing, and the domino effect of people leaning against each other.

“Compressive asphyxia has occurred from people being stacked up vertically, one on top of the other, or horizontal pushing and leaning forces. In the Ibrox Park soccer stadium incident, police reported that the pile of bodies was 3 m (10 feet) high. At this height, people on the bottom would experience chest pressures of 3600-4000 N (800-900 lbs.), assuming half the weight of those above was concentrated in the upper body area.

“Horizontal forces sufficient to cause compressive asphyxia would be more dynamic as people push off against each other to obtain breathing space. In the Cincinnati rock concert incident, a line of bodies was found approximately 9 m (30 ft) from a wall near the entrance. This indicates that crowd pressures probably came from both directions as rear ranks pressed forward and front ranks pushed off the wall.

“Experiments to determine concentrated forces on guardrails due to leaning and pushing have shown that force of 30% to 75% of participant weight can occur. In a US National Bureau of Standards study of guardrails, three persons exerted a leaning force of 792 N (178 lbs.) and 609 N (137 lbs.) pushing. [9] In a similar Australian Building Technology Centre study, three persons in a combined leaning an pushing posture developed a force of 1370 N (306 lbs.). [10] This study showed that under a simulated "panic", 5 persons were capable of developing a force of 3430 N (766 lbs.).

[9] Fattal, S.G., Cattaneo, L.E. Investigation of Guardrails for the Protection of Employees From Occupational Hazards. Nat. Bur. Stds. NBSIR 76-1139, July 1976, 114 pp.

[10] Horizontal Loading on Handrails. NBTC Tech. Rec 514, Nat.Tech. Centre, New South Wales.

6. 20% compression of the chest cavity can break ribs - this is around 3000N (from medical bone fracture analysis) so we can see that there is a potential for a 5 person deep push to break ribs. 

is from “Biomechanics of Chest and Abdomen Impact” (David C. Viano, Albert I King) 2000 CRC Press LLC (which I’ve attached)

7. Gross Bone Fractures (Pathological and physical parameters) are around 6920N - which is above the concert and football barrier tolerances (5400N)

is from Patellofemoral Joint Fracture Load Prediction using Physical and Pathalogical Parametres” Proceedings of the 1998 SAE International Congress and Exposition. SAE 980358 by PJ Atkinson, CM Mackenzie and RC Haunt.


Prediction of Human Crowd Pressures (Ris S.C. Lee, Roger Hughes) Accident Analysis and Prevention 2006 Elsevier. Is a collection of the above on one paper.

© G. Keith Still.  All rights reserved. All images, videos, audio files and text are protected under international copyright law  You may not copy, modify or use any of the website content, in whole or in part, for any commercial or non-commercial purpose without permission.  Contact email