Hubbry Logo
Crowd collapses and crushesCrowd collapses and crushesMain
Open search
Crowd collapses and crushes
Community hub
Crowd collapses and crushes
logo
8 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Crowd collapses and crushes
Crowd collapses and crushes
Crowd collapses and crushes
View on Wikipedia
from Wikipedia

Black-and-white engraving depicting 1711 human crush on an arch bridge
245 people died in the Lyon bridge disaster of 1711, when a large crowd returning from a festival on one side of the bridge found their way blocked by a collision between a carriage and a cart, and became trapped.

Crowd collapses and crowd crushes are catastrophic incidents that occur when a body of people becomes dangerously overcrowded. When numbers are up to about five people per square meter,[a] the environment may feel cramped but manageable; when numbers reach between eight and ten people per square meter,[1][b] individuals become pressed against each other and may be swept along against their will by the motion of the crowd.[2] Under these conditions, the crowd may undergo a progressive collapse where the pressure pushes people off their feet, resulting in people being trampled or crushed by the weight of other people falling on top of them. At even higher densities, the pressure on each individual can cause them to be crushed or asphyxiated while still upright.[3]

Such incidents are invariably the product of organizational failures, and most major crowd disasters could have been prevented by simple crowd management strategies.[4] Such incidents can occur at large gatherings such as sporting, commercial, social, and religious events. The critical factor is crowd density rather than crowd size.[5]

Crowd collapses and crushes are often reported incorrectly as human stampedes, which typically occur when a large group of people all try to get away from a perceived risk to life.[6]

Background

[edit]

One study has calculated that there were 232 deaths and over 65,000 injuries in the ten years between 1992 and 2002 as a result of such incidents,[7] but crowd scientists believe that such casualties are both vastly under-reported and increasing in frequency. One estimate is that only one in ten crowd injuries occurring in doorbuster sales are reported, while many, if not most, injuries at rock concerts go unreported.[7]

Dynamics

[edit]
Main article: Crowd analysis

The average individual occupies an oval floorspace approximately 30 by 60 centimetres (1 by 2 ft)—0.2 square metres; 2 square feet—and at densities of 1 to 2 per square meter (or per 10 ft2) individuals can move freely without contact.[3] Even if people are moving quickly, at this density one can avoid obstacles, and the chance of a crowd-related incident is minimal. Even at three or four people per square meter,[c] the risk is low;[8] however, at densities of five per square meter,[a] it becomes more difficult for individuals to move, and at higher densities of six to seven per square meter,[d] individuals become pressed against each other and can be unable to move voluntarily. At this point a crowd can begin to behave like a fluid, with individuals moved about by the pressure of those around them, and shockwaves can pass through the crowd as pressures within the crowd change.[9][3] This can be highly dangerous, although some people actively seek this experience, such as at rock concerts[10] or football matches,[11] where the excitement, camaraderie, and literally "going with the flow" is for some an essential part of the experience,[12] and activities like dancing and moshing are common. The danger inherent in these conditions is that the crowd will collapse in on itself or become so densely packed that individuals are crushed and asphyxiated.

Crowd collapses

[edit]

A crowd collapse occurs when a crowd is so dense that each individual is touching others all around and is, to an extent, supported by those around. This can occur whether the crowd is moving or stationary. If a person then falls, the support to those around is lost, while the pressure from those further out remains, causing people to fall into the void. This process is then repeated, causing a bigger void, and will progress until the pressure eases; meanwhile, those who have fallen are at risk of being smothered by the weight of bodies on top or being trampled as the crowd is swept over them.[5] An example of a progressive crowd collapse was the 2015 Mina stampede in Mecca, Saudi Arabia during the Hajj[13] when over 2,400 people were reported to have died.

Crowd crushes

[edit]

At even higher densities (approaching ten people per square metre (one per square foot))[1] a crowd can become so packed that people are crushed together to such an extent they can no longer breathe and are asphyxiated.[3] Such crowd crushes can occur when a moving crowd is funneled into a smaller and smaller space, when it meets an obstacle (such as a dead end, or a locked door), or when an already densely packed crowd has an influx of people, causing a pressure wave toward those at the front of the crowd. In this situation those entering may be unaware of the effect on those in front and continue to press in.[5] Examples of crushes are the Hillsborough disaster in Sheffield, South Yorkshire, England in 1989, the Love Parade disaster in Duisburg, North Rhine-Westphalia, Germany in 2010, the Astroworld Festival crowd crush in Houston, Texas, in 2021, and the Itaewon Halloween crowd crush in Itaewon, Seoul, South Korea in 2022.[13]

Crowd "stampedes"

[edit]

Stampede is a loaded word as it apportions blame to the victims for behaving in an irrational, self-destructive, unthinking and uncaring manner, it's pure ignorance, and laziness ... It gives the impression that it was a mindless crowd only caring about themselves, and they were prepared to crush people. In virtually all situations it is usually the authorities to blame for poor planning, poor design, poor control, poor policing and mismanagement.

Edwin Galea, professor of fire safety engineering at the University of Greenwich, England[14]
Main article: Stampede § Human stampedes and crushes

The term "stampede" is usually used in reference to animals that are fleeing a threat. Stampede events that involve humans are extremely rare and are unlikely to be fatal.[5] According to Keith Still, professor of crowd science at Manchester Metropolitan University, "If you look at the analysis, I've not seen any instances of the cause of mass fatalities being a stampede. People don't die because they panic. They panic because they are dying".[5] Paul Torrens, a professor at the Center for Geospatial Information Science at the University of Maryland, remarks that "the idea of the hysterical mass is a myth".[5] Incidents involving crowds are often reported by media as the results of panic.[15][1] However, the scientific literature has explained how panic is a myth which is used to mislead the attention of the public from the real causes of crowd incidents, such as a crowd crush.[16][17][18]

Causes of death

[edit]

In crowd collapse and crush incidents the most common cause of death is asphyxiation, caused either by vertical stacking, as people fall on top of one another, or by horizontal stacking, where people are crushed together or against an unyielding barrier. Victims can also exhibit bone fractures caused by the pressure,[19] or trampling injuries, when a crowd has swept over them where they have lain.[19]

Prevention

[edit]
Main article: Crowd control

It is believed that most major crowd disasters can be prevented by simple crowd management strategies.[20] Crushes can be prevented by organization and traffic control, such as crowd barriers. On the other hand, barriers in some cases may funnel the crowd toward an already-packed area, such as in the Hillsborough disaster. Hence barriers can be a solution in preventing or a key factor in causing a crush. One problem is lack of feedback from people being crushed to the crowd pressing behind—feedback can instead be provided by police, organizers, or other observers, particularly raised observers, such as on platforms or horseback, who can survey the crowd and use loudspeakers to communicate and direct a crowd.[21] In some cases it may be possible to take simple measures such as spreading movements out over time.[22]

A factor that may contribute to a crush is inexperienced security officers who assume that people's behaviour in a dense crowd is voluntary and dangerous, and start applying force or preventing people from moving in certain directions. In the 1989 Hillsborough disaster, some police and stewards were so concerned with what they saw as possible hooliganism that they took actions that actually made matters worse.[22]

There is risk of a crush when crowd density exceeds about five people per square meter.[a] For a person in a crowd a signal of danger, and a warning to get out of the crowd if possible, is the sensation of being touched on all four sides. A later, more serious, warning is when one feels shock waves travelling through the crowd, due to people at the back pushing forward against people at the front with nowhere to go.[21] Keith Still of the Fire Safety Engineering Group, University of Greenwich, said "Be aware of your surroundings. Look ahead. Listen to the crowd noise. If you start finding yourself in a crowd surge, wait for the surge to come, go with it, and move sideways. Keep moving with it and sideways, with it and sideways."[5] Other recommendations include trying to remain upright, and keeping away from walls and other obstructions if possible.[23]

After the 1883 crush known as the Victoria Hall disaster in Sunderland, England, which killed 183 children, a law was passed in England which required all public entertainment venues to be equipped with doors that open outwards—for example, using crash bar latches that open when pushed.[24] Crash bars are required by various building codes.

See also

[edit]

Notes

[edit]

References

[edit]

Sources

[edit]
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

Crowd collapses and crushes

View on Grokipedia
from Grokipedia
Crowd collapses and crushes are physical disasters in densely packed assemblies where interpersonal forces generate compressive pressures exceeding human tolerance, resulting in injuries and fatalities chiefly from traumatic asphyxia rather than trampling or panic-driven flight. These events arise when crowd densities surpass critical thresholds—typically above 4 to 6 individuals per square meter—triggering instabilities akin to fluid turbulence, including stop-and-go waves and pressure buildups that propagate uncontrollably. Empirical analyses of such incidents reveal that causal factors stem from bottlenecks, converging flows, or ingress/egress imbalances, independent of deliberate panic in many cases, challenging outdated narratives of irrational "stampede" behavior. Defining characteristics include the predominance of crush asphyxia over other trauma, with prevention rooted in engineering controls like density monitoring and flow optimization rather than solely behavioral interventions. Notable scholarly advancements, such as models of crowd dynamics, underscore how simple local interactions yield macroscopic disasters, informing risk assessment in mass gatherings like religious pilgrimages or events. Controversies persist over terminological precision, with "collapse" denoting structural or crowd failure under load and "crush" emphasizing sustained compression, yet both highlight vulnerabilities in unmanaged high-density environments.

Terminology and Conceptual Framework

Definitions and Distinctions

Crowd crushes occur when high-density assemblies generate sustained lateral compressive forces that thoracic expansion, leading to as individuals cannot inhale sufficiently against the exerted by surrounding bodies or fixed barriers. This physiological , known as compressive , predominates in such incidents, for nearly all fatalities rather than injuries from or blunt . Densities exceeding 7 persons per square meter mark a critical threshold where crowds transition to fluid-like , amplifying forces to levels—often thousands of newtons per meter squared—that exceed postural stability and respiratory capacity. Crowd collapses, by contrast, involve a vertical dimension wherein an initiating fall or imbalance in a compressed crowd propagates sequentially, causing adjacent individuals to topple and form layered piles that intensify localized compression and obstruct escape or ventilation. Termed progressive crowd collapse, this mechanism resembles a domino effect in ultra-dense conditions (>6 persons per square meter), where reliance on mutual support fails, resulting in rapid body stacking and secondary asphyxiation. The distinction hinges on force orientation: crushes emphasize horizontal squeezing from crowd momentum or confinement, often without initial vertical displacement, while collapses incorporate gravitational piling that exacerbates horizontal pressures through reduced effective density in voids. Both phenomena stem from biomechanical overload rather than behavioral frenzy, with empirical analyses of disasters confirming asphyxia via autopsy over trauma.

Debunking Panic and Stampede Myths

A persistent misconception attributes crowd collapses and crushes to "mass panic" or "stampede," depicting crowds as irrational, self-destructive entities that trample victims in frenzied flight. This narrative, prevalent in media coverage, implies victim culpability through chaotic behavior, thereby deflecting scrutiny from infrastructural and managerial shortcomings. Empirical reviews of disaster autopsies and eyewitness data contradict this view, showing panic to be exceptional rather than normative. Civil engineer John J. Fruin, analyzing over 20 major incidents including the 1971 Ibrox Park collapse (66 deaths) and 1970 Cincinnati Riverfront Coliseum crush (1 death, multiple injuries), determined that compressive asphyxiation—arising from sustained chest pressures exceeding 3,600–4,000 Newtons—accounts for virtually all fatalities, not trampling or falls amid running. In the 2010 Love Parade disaster (21 deaths), medical examiners confirmed suffocation and back injuries from overcrowding, with no evidence of widespread trampling; press characterizations of "stampede" misaligned with on-site findings of density-induced stasis. Crowd physicist Dirk Helbing further delineates "crowd quakes"—progressive collapses triggered by density waves or minor perturbations, as in the 2006 Mina Hajj incident (345+ deaths)—from purported panic events, emphasizing that such quakes propagate without requiring hysterical flight. Video analyses and simulations reveal participants often cooperating to aid the fallen, only succumbing to involuntary compression at densities above 4 persons per square meter, where diaphragmatic restriction prevents breathing. Professor Edwin Galea encapsulates this reversal: "People don’t die because they panic. They panic because they are dying," underscoring that distress signals emerge from physiological peril, not antecedent hysteria. Perpetuating the panic myth fosters inadequate countermeasures, such as overemphasizing behavioral controls over capacity limits and egress design, as evidenced by recurring failures in narrow venues like Seoul's 2022 Itaewon alley (158+ deaths from unchecked inflow). Correcting this requires recognizing crowd dynamics as governed by physical constraints—akin to granular flow—wherein rational actors yield to biomechanical inevitabilities at critical thresholds, absent any "hysterical mass" archetype.

Physical and Behavioral Mechanics

Biomechanical Forces and Density Limits

In dense crowds, biomechanical forces arise primarily from interpersonal contacts and compressive pressures that exceed human postural stability and respiratory tolerances. At densities below approximately 2-3 persons per square meter, individuals maintain sufficient personal space for voluntary movement and balance, with minimal sustained contact forces. However, as density increases to 4-5 persons per square meter, contact forces become frequent, reducing walking speeds and introducing involuntary pushing, where average forces per contact can reach 100-200 N, leading to discomfort and reduced maneuverability. Critical density thresholds for collapse and crush typically occur at 6-7 persons per square meter, where crowds transition to fluid-like dynamics, propagating shock waves from minor perturbations that destabilize balance across multiple individuals. At this level, the effective personal space shrinks to under 0.17 square meters, preventing arm usage for self-protection or equilibrium recovery, and generating sustained compressive forces up to 400 N per person if pressure builds for over 30 seconds, causing thoracic strain and potential falls in chain reactions known as progressive collapses. Empirical models from crowd flow experiments confirm that flow capacity peaks around 4 persons per square meter before declining sharply, with turbulence-like instabilities emerging above 6 persons per square meter, analogous to jamming transitions in granular materials but driven by human biomechanics. Beyond these limits, at 8-10 persons per square meter observed in disasters like the 2010 Love Parade incident, crowds become quasi-static masses incapable of directed motion, with vertical and horizontal forces amplifying through piling or sustained compression. Biomechanical analysis indicates that fatal compression asphyxia requires acute chest forces of at least 500-1000 N to induce flail chest or diaphragmatic paralysis, often via rib fractures, as modeled from cadaveric and historical data; in crowds, these emerge from the cumulative weight of overlying bodies (approximately 50-100 kg per layer) restricting thoracic expansion and venous return, leading to rapid hypoxia within 3-6 minutes without relief. Such forces are not uniformly distributed but localize on vulnerable anterior chests, exacerbated by inability to brace, with peer-reviewed thresholds confirming that pressures exceeding 20-30 kPa on the torso suffice for respiratory arrest in prone or supine positions. Density limits are further constrained by human anthropometrics: average adult shoulder width (0.4-0.5 m) and depth (0.3 m) imply a theoretical maximum packing of 5-7 persons per square meter before overlap forces ribs and abdomen into immobility, corroborated by discrete element simulations showing stress concentrations that predict crush zones. Risk escalates nonlinearly due to heterogeneity in crowd composition—e.g., children or elderly have lower tolerance thresholds, amplifying local forces by 20-50% during surges. These biomechanical realities underscore that exceeding density limits inherently generates uncontrollable forces, independent of behavioral panic, as evidenced by empirical data from controlled experiments and post-incident analyses.

Flow Dynamics and Trigger Mechanisms

Crowd flow dynamics in dense assemblies resemble fluid mechanics, where pedestrian movement generates kinematic waves that propagate upstream as stop-and-go patterns, leading to density fluctuations even without external perturbations. These waves arise from inertial effects and local interactions, with propagation speeds inversely related to density; at low densities (below 2 persons per square meter), flows remain laminar, but above 4 persons per square meter, oscillatory instabilities emerge, amplifying pressure gradients. Empirical measurements from controlled experiments confirm that bidirectional or converging flows on ramps or narrow paths accelerate wave formation, reducing overall throughput by up to 50% compared to unidirectional movement. At critical densities exceeding 6 persons per square meter, collective motion transitions to compressive asphyxia risks, as body contact forces exceed voluntary control, enforcing piston-like surges where rearward individuals propel forward ones uncontrollably. Physics-based models, such as those extending kinematic wave theory to pedestrian streams, quantify this by relating flow rate qq to density ρ\rho via q=ρv(ρ)q = \rho v(\rho), where velocity vv drops sharply beyond 4-5 persons per square meter, fostering "turbulent" bursts of disorder akin to pressure eruptions in granular media. In such regimes, crowds self-organize into chiral oscillations or density hotspots, as observed in large-scale data from events involving thousands, where orbital motions coordinate hundreds without centralized cues. Trigger mechanisms typically stem from spatial constraints rather than behavioral panic, initiating when inflow rates surpass egress capacities at bottlenecks, compressing crowds into densities over 10 persons per square meter. For instance, in the 2010 Love Parade disaster, a narrow tunnel entrance created a convergence point where upstream momentum from 1,000+ entrants per minute overwhelmed the 600-person safe limit, generating compressive forces without initial fleeing. Similarly, bidirectional streams, as in the 2022 Seoul Halloween crush, provoke counterflows that halt progress and build lateral pressures, with empirical video analysis showing initial triggers from minor obstructions amplifying into full collapses via chain-reaction falls. Structural or informational asymmetries—such as uneven path widths or unheeded capacity signals—further catalyze these by inducing asymmetric surges, where faster-moving subgroups overrun slower ones ahead. High-density empirical studies underscore that such triggers operate mechanically, with falling individuals creating trip hazards that cascade into pile-ups, independent of emotional states.

Causal Factors and Risk Analysis

Organizational and Infrastructure Deficiencies

Organizational deficiencies contributing to crowd collapses and crushes typically involve failures in pre-event risk assessment, capacity planning, and inter-agency coordination. In the 1989 Hillsborough Stadium disaster, police command decisions prioritized hooliganism containment over ingress monitoring, resulting in the opening of a large gate that funneled approximately 2,000 fans into already crowded terrace pens holding up to 5,000, without adjusting stewarding or halting the match. The Taylor Interim Report attributed this to inadequate planning and over-reliance on outdated crowd control tactics, exacerbating density beyond safe limits in the central pens. Similarly, the 2010 Love Parade in Duisburg suffered from organizers' underestimation of attendance—forecasting 1.4 million but capping permits at 250,000—coupled with delayed event start times due to site preparation, which compressed arrival flows without adaptive controls. Insufficient staffing and communication breakdowns compound these issues by impairing real-time response. At Hillsborough, radio failures and mismatched command structures delayed recognition of the crush, with police logs showing no coordinated halt to inflows despite visible overcrowding at 2:50 p.m. In Duisburg, over 3,000 police officers were deployed but operated without unified loudspeakers or flow directives, leading to abandoned entry cordons as crowds pressured barriers; a shift change further disrupted oversight during peak density. Official analyses frame these as systemic lapses, where contingency protocols—such as scalable entry throttling—were either absent or overridden by pressure to maintain event momentum, ignoring prior overcrowding signals from 1988 at the same venue. Infrastructure deficiencies often manifest as physical constraints that amplify organizational errors through bottlenecks and flawed layouts. The Love Parade site's primary access via Karl-Lehr-Straße tunnel, functioning as both entry and exit without segregated flows, created a funnel effect; fencing reduced the adjacent ramp width to 10.59 meters, capping throughput at roughly 52,000 persons per hour against unmanaged surges exceeding critical densities of 4-5 persons per square meter. Hillsborough's Leppings Lane end featured a steep, radial-fanned tunnel directly feeding into terraced pens, promoting surging without intermediate buffers, while radial perimeter fences—designed for containment rather than egress—trapped fans during the 3:00 p.m. kickoff crush, contributing to barrier failures under compressive loads. These design relics, unadapted for post-1960s crowd volumes, violated basic flow principles by lacking wave-breaking elements like columns or widened merges, as evidenced in post-incident force reconstructions showing rail deflections over 4,500 Newtons. In recurrent mass gatherings like the Hajj, fixed road infrastructure—such as narrow Jamarat Bridge approaches—interacts with organizational oversights in pilgrim routing, where unmanaged surges from adjacent ritual sites have repeatedly exceeded egress capacities, as seen in the 2006 Mina incident with 363 deaths amid 750,000 participants funneled through hazardous eastern entrances lacking surge mitigation. Empirical patterns from such events underscore that unaddressed venue flaws, including incomplete facilities or obstacle-cluttered paths (e.g., vehicles on Love Parade ramps), propagate instabilities like crowd turbulence, independent of behavioral panic myths.

Environmental and Human Behavioral Contributors

Extreme heat during mass gatherings can induce dehydration and heat-related illnesses, reducing individuals' physical resilience against crowd pressures and increasing susceptibility to collapse. Studies of outdoor events report heat stress as a leading environmental hazard, contributing to fatigue and impaired coordination in dense crowds. Similarly, heavy rain creates slippery surfaces, elevating the risk of slips and prompting sudden surges toward shelter, which can initiate compressive forces. Terrain irregularities, such as slopes or stairs, exacerbate instability by facilitating loss of footing and chain-reaction falls. In the 2022 Itaewon Halloween crowd crush in Seoul, a sloped alleyway contributed to a domino effect, with witnesses describing crowds losing balance on the incline, leading to 159 deaths. Research on evacuation dynamics confirms that steeper slopes amplify injury severity and propagation of falls in high-density scenarios, with slopes exceeding 4° notably hindering smooth movement. Narrow corners and uneven topography, as in 68.42% of organized stampede cases involving stairs, heighten tripping risks and congestion. Human behavioral factors often involve competitive tendencies and inadequate spatial awareness, driving convergence at bottlenecks or focal points rather than disorderly flight. Analysis of 209 overcrowding incidents from 2000–2022 identified competitive psychology in 65.22% of spontaneous gatherings, prompting pushing that builds unsafe densities. Falling or squatting, observed in up to 78.95% of applied activity cases, acts as a primary trigger by obstructing flow and compressing downstream individuals. Weak hazard awareness and emergency response skills, prevalent in 68.42% of applied events, lead to delayed or ineffective dispersal, as evidenced by poor public safety concepts in over 70% of organized cases. In religious or celebratory contexts, behavioral attraction to specific sites, such as relics or stages, causes localized surges independent of panic, with reviews of incidents like the Kumbh Mela highlighting group dynamics over individual fear as density amplifiers. These patterns underscore that behavioral contributors typically stem from affiliative or competitive motivations rather than irrational terror, aligning with empirical observations that crowds maintain cooperative elements until critical densities are reached.

Incident Typology

Progressive Collapses

Progressive crowd collapses occur when localized instability in a densely packed crowd—such as one or more individuals falling due to excessive lateral pressure, tripping, or slipping—triggers a chain reaction of subsequent falls, resulting in a propagating wave of people toppling into the void and forming human piles that crush those at the base through compressive asphyxiation or traumatic injuries. This mechanism differs from sustained compressive crushes, where victims remain upright under prolonged static pressure without widespread falling, as the progressive variant involves dynamic propagation akin to a domino effect, amplifying the collapse front across the crowd. Crowd densities exceeding 6-7 persons per square meter often precede such events, where individuals lose independent mobility and are carried by collective forces, with early indicators including people being lifted off their feet and shuffled forward involuntarily. The initiation typically stems from transient surges in a moving or funneling crowd, such as through bottlenecks or during radial flows toward a focal point, where forward momentum overrides stability; once the initial fall creates a gap, adjacent individuals tumble in due to unbalanced forces, enlarging the affected area and escalating the pile height, which can reach multiple layers and exceed lethal compressive loads of 300-500 kg per square meter on lower victims. Biomechanical models simulate this as a failure propagation in granular-like media, with standing spectators in terraced venues particularly vulnerable due to uneven terrain and persistent pressure from behind, as analyzed in football stadium contexts where modeling tools predict collapse thresholds based on occupancy and standing density. Unlike panic-driven myths, these collapses arise from physical overcrowding rather than irrational flight, with empirical data showing no correlation to fear but strong links to unmanaged inflow exceeding egress capacity. Notable incidents illustrate this typology, including the 2015 Mina stampede during Hajj pilgrimage on September 24, where over 2,200 fatalities resulted from progressive collapse amid 2-3 million pilgrims converging on a narrowing route, with autopsy evidence confirming layered piling and suffocation from sequential falls rather than isolated trampling. Similarly, in the 1989 Hillsborough disaster on April 15, initial perimeter fencing and police-directed surges into overcrowded pens initiated falls that propagated into a multi-tiered crush, killing 97 via compressive forces on fallen individuals, as detailed in subsequent engineering inquiries emphasizing density gradients over behavioral panic. These cases underscore that progressive collapses amplify mortality through vertical stacking, with survival rates dropping below 10% for those buried beyond the second layer due to impeded respiration and circulation.

Compressive Crushes

Compressive crushes represent a subtype of crowd disaster characterized by sustained lateral compression in high-density assemblies, where individuals experience prolonged pressure against adjacent bodies or fixed obstacles, primarily resulting in asphyxia rather than dynamic toppling or trampling. These incidents typically manifest in relatively static or low-mobility crowds, such as those converging at bottlenecks or accumulating in confined spaces without rapid evacuation, leading to densities that generate immobilizing forces. Unlike surges initiating falls, compressive dynamics emphasize bidirectional or multidirectional pressures that maintain upright postures while restricting thoracic expansion. The biomechanical threshold for harm arises at crowd densities exceeding 10 persons per square meter, beyond which compressive forces impair mobility and respiration; critical levels of 15-20 persons per square meter or higher correlate with widespread asphyxia, as simulated in physics-based models using rigid-body agents under controlled compression. Forces in such scenarios can reach 4450 N per meter of frontage, equivalent to the output of 6-7 adults pushing unidirectionally, sufficient to deform steel barriers and prevent diaphragm function at sustained pressures above 3-5 kPa. Pathophysiologically, victims suffer positional asphyxia, with chest compression halting ventilatory efforts within 1-3 minutes, followed by hypercapnia and cardiac arrest; autopsies from such events confirm traumatic asphyxia as the dominant mechanism, accounting for the majority of fatalities over blunt injuries. Empirical data indicate compressive crushes predominate in unmanaged entry points or ritual gatherings, where inflow rates outpace egress or spatial capacity, as evidenced by simulations showing 25% area reduction elevating densities to injurious levels for over 80% of participants. Safe static densities remain below 5 persons per square meter to avert force buildup, with monitoring tools like density mapping essential for early intervention. In contrast to progressive collapses driven by inertial surges and sequential falls, compressive variants sustain pressure gradients without widespread collapse, amplifying lethality through immobility and secondary organ failure from prolonged hypoxia.

Mortality and Injury Profiles

Primary Pathophysiological Causes

The predominant cause of mortality in crowd crushes is traumatic asphyxia, characterized by sustained compressive forces on the thorax and upper abdomen that mechanically restrict respiratory excursions, leading to hypoventilation, hypercapnia, hypoxia, and rapid cardiorespiratory arrest. This occurs when crowd densities exceed 6-7 individuals per square meter, generating horizontal forces sufficient to immobilize victims and prevent diaphragmatic descent or rib cage expansion, often within 30 seconds of onset, with unconsciousness following shortly thereafter and death ensuing in approximately 6 minutes without relief. Forensic analyses of incidents, including the 2022 Itaewon crush, confirm that compressive asphyxia accounts for the majority of fatalities, with autopsy findings revealing petechial hemorrhages in the face and neck, pulmonary edema, and absence of significant skeletal trauma as hallmarks, distinguishing it from blunt force mechanisms. Secondary pathophysiological processes exacerbate outcomes but are not primary drivers in acute crushes; these include positional asphyxia from body stacking or prone positioning under pressure, which further compromises venous return and oxygenation, and rare instances of crush syndrome involving rhabdomyolysis and electrolyte derangements from prolonged limb compression, though the latter predominates in scenarios like structural collapses rather than dynamic crowd dynamics. Empirical data from historical crowd disasters indicate that trampling-related injuries, such as lower extremity fractures or head trauma, contribute to morbidity but rarely to mortality, as immobilized victims in dense compressions lack the mobility for widespread trampling to occur. Vulnerable populations, including children, the elderly, and those with preexisting cardiopulmonary conditions, exhibit accelerated decompensation due to reduced respiratory reserve, with forces as low as 3-4 kN/m² sufficient to induce failure in these groups. In compressive asphyxia, the pathophysiological cascade begins with mechanical impedance of tidal volume, progressing to alveolar hypoventilation and arterial desaturation; elevated intrathoracic pressure impairs cardiac output via reduced preload, culminating in metabolic acidosis and multiorgan failure if compression persists beyond 5-10 minutes. Post-mortem studies underscore that survival hinges on early decompression, as even brief episodes can cause irreversible cerebral hypoxia, with no effective field interventions beyond physical extraction. This mechanism aligns with biomechanical models of crowd flow, where anisotropic pressure gradients—higher anteriorly—selectively target the chest, explaining the positional invariance of fatalities across incident typologies.

Empirical Patterns from Data

A comprehensive review of 215 human stampede events reported between 1980 and 2007 documented 7,069 fatalities and at least 14,078 injuries across 213 incidents with available data, yielding an average of approximately 33 deaths and 66 injuries per event. Fatality rates per 100,000 attendees varied widely: 23.3% of events recorded zero deaths, 35.3% had rates below 100 per 100,000, and 41.4% exceeded 100 per 100,000, with higher rates correlating to denser crowd configurations and bottlenecks. Injuries outnumbered fatalities by a factor of roughly 2:1 in this dataset, predominantly comprising musculoskeletal trauma, abrasions, and concussions, though systematic underreporting of non-fatal injuries likely understates the ratio. Forensic and epidemiological analyses consistently identify compressive asphyxia—resulting from sustained thoracic compression impairing diaphragmatic excursion and ventilation—as the proximate cause in nearly all crowd crush fatalities, rather than traumatic injuries like trampling or blunt force. Autopsy data from multiple incidents reveal that victims succumb within 5-10 minutes of onset due to hypercapnia and hypoxia, with post-mortem evidence including petechial hemorrhages, pulmonary edema, and absence of significant skeletal fractures in asphyxia-dominant cases. A tensor decomposition of 186 fatal crowd accidents from 1979 to 2023 found that events driven solely by internal crowd dynamics (e.g., surges without external triggers) exhibited elevated mortality rates of 741 per 100,000 exposed individuals, compared to 567 per 100,000 in incidents involving exogenous factors like structural failures. Aggregate data from 123 documented fatal human crushes between 1900 and 2017 indicate a median of 15-20 deaths per incident, with outliers in mass religious gatherings exceeding 1,000 fatalities due to prolonged compression in confined spaces. Injury profiles skew toward non-lethal compressive and shear forces, with reported cases showing 2-4 times more survivors experiencing rib fractures, soft tissue contusions, and respiratory distress than those succumbing to terminal asphyxia. Temporal trends from press and media compilations (1900-2019) reveal no significant decline in per-event casualty severity despite population growth, suggesting persistent vulnerabilities in density thresholds above 4-6 persons per square meter. Demographic patterns highlight elevated risks for children and smaller-statured individuals, who comprise disproportionate shares of asphyxia victims owing to positional suffocation under stacked bodies.

Mitigation and Prevention Approaches

Engineering and Venue Design Principles

Venue design principles for mitigating crowd collapses and crushes prioritize structural resilience, controlled densities, and efficient circulation to prevent compressive forces and progressive failures. Safe capacity assessments form the foundation, limiting occupant loads to levels that avoid overcrowding while ensuring adequate space for movement; for instance, the UK's Guide to Safety at Sports Grounds (Green Guide) mandates calculations by competent professionals, incorporating viewing standards, holding capacities, and egress flows, with standing densities capped at approximately 4.7 persons per square meter in controlled viewing areas to minimize compression risks. Similarly, NFPA 101 for assembly occupancies requires scaling egress provisions to occupant loads, mandating at least two means of egress for up to 500 persons and additional exits thereafter, with corridor widths of at least 44 inches for 50 or more occupants to facilitate unimpeded evacuation. Structural engineering incorporates live load specifications tailored to crowd dynamics, distinguishing static weights from induced vibrations. International codes, such as Eurocode 1 and equivalents in BS 6399 or AS/NZS 1170, prescribe assembly area loads of 4-5 kN/m², augmented by dynamic factors for rhythmic activities like jumping, where amplification coefficients can reach 1.8 or higher under resonance, necessitating vibration analysis and damping elements in floors, balconies, and raker beams to avert fatigue or collapse. Balustrades and barriers must resist horizontal crowd pressures, often designed for 3-5 kN/m line loads in high-density zones, with research emphasizing slenderness limits to prevent toppling or failure propagation. Circulation and layout features emphasize bottleneck avoidance through wide, unobstructed paths and flow-guiding elements. Egress components, per NFPA 101, demand clear widths scaled to loads—typically 0.15-0.2 inches per person for level paths and stairs—while prohibiting dead-end corridors exceeding 20 feet in certain assemblies and ensuring doors swing in the direction of egress travel for loads over 50. Engineered barriers, including retractable stanchions or fixed railings, segment crowds into subgroups and control ingress/egress rates, with designs tested for surge resistance to distribute forces and maintain separation, as demonstrated in post-incident analyses where inadequate partitioning exacerbated crushes. Integration of simulation tools in design verifies these principles, modeling crowd flows to identify pinch points and validate capacities under peak loads. Dynamic modeling, aligned with Green Guide recommendations, simulates densities and velocities to refine layouts, ensuring entry/exit rates match holding capacities and reducing vulnerability to density spikes that precipitate collapses. Overall, these approaches, grounded in empirical load data and code-mandated factors, shift focus from reactive panic assumptions to proactive density and force management, though implementation varies by jurisdiction and requires ongoing validation against real-world dynamics.

Operational and Regulatory Strategies

Operational strategies for preventing crowd collapses and crushes emphasize proactive planning, real-time monitoring, and adaptive flow control to maintain densities below critical thresholds where compressive forces can lead to asphyxia or structural failure. Event organizers typically conduct density mapping and ingress/egress simulations using tools like crowd dynamics software to identify pinch points, with capacities capped at 1.5-2 persons per square meter in standing areas to avoid progressive collapses. Stewards and security personnel are deployed at ratios of 1 per 50-100 attendees, trained in de-escalation and evacuation protocols, including the use of barriers to segment crowds into manageable zones and sliding gates for controlled entry to prevent surges. Real-time surveillance via CCTV and mobile apps enables dynamic adjustments, such as halting inflows when upstream densities exceed safe limits, as demonstrated in post-incident analyses of events like the 2010 Love Parade where inadequate monitoring contributed to 21 fatalities. Training programs for staff focus on recognizing behavioral precursors to crushes, such as bidirectional flows or audience surges toward stages, with protocols to widen sight lines and distribute viewing areas evenly to minimize forward pressure. Emergency response integration includes on-site medical teams equipped for crush syndrome treatment, with communication systems linking organizers, authorities, and attendees via apps for alerts on overcrowding. Post-event debriefs incorporate data from sensors measuring footfall and pressure to refine future operations, prioritizing empirical metrics over anecdotal estimates. Regulatory frameworks enforce these strategies through mandatory compliance with standards like NFPA 101 Life Safety Code, which specifies egress widths of at least 0.2 meters per occupant in assembly spaces and requires fire-rated separations to contain crowd movements. In the United States, OSHA mandates pre-event risk assessments for retail or public gatherings, including barricades at least 10 feet from entry points and trained crowd managers numbering one per 250 attendees outdoors, with violations punishable by fines up to $14,502 per instance as of 2023 adjustments. Jurisdictions like the UK require event permits under the Health and Safety at Work Act 1974, mandating hydraulic modeling for crowd flows and independent audits, while international bodies such as the UN's Sendai Framework advocate for building codes incorporating dynamic load factors up to 5 kN/m² for temporary structures. Enforcement often involves licensing authorities rejecting plans exceeding venue capacities by more than 10%, with post-event investigations by bodies like the UK's HSE leading to prosecutions, as in the 2021 Hawthorn Woods festival case where overcrowding breached density regulations. These regulations prioritize verifiable engineering data over self-reported attendance, countering underestimation biases in organizer projections, and increasingly incorporate AI-driven predictive analytics for permit approvals to flag high-risk configurations. Non-compliance has prompted updates, such as Australia's 2023 crowd safety laws post-festival incidents, requiring real-time density reporting and penalties scaled to event size.

Empirical Case Studies

Pre-Modern and Early Incidents

One of the earliest documented cases of a crowd crush occurred on October 11, 1711, on the Pont de la Guillotière spanning the Rhône River in Lyon, France. A dense crowd of revelers returning from a religious festival across the river encountered gates closed by a local merchant to shield his sheep from the approaching throng. This obstruction triggered compressive asphyxia and falls into the water, resulting in hundreds of fatalities amid the ensuing panic. The merchant's wife, Catherine Mazenod, was implicated in the decision, leading to her subsequent donation of substantial lands, known as Part-Dieu, to the city's charitable institutions as restitution. Such pre-modern incidents often stemmed from rudimentary infrastructure unable to handle surge densities and abrupt flow interruptions, exemplifying causal dynamics of crowd compression independent of modern panic myths. Limited contemporary records constrain precise mortality figures, with estimates varying widely due to inconsistent reporting practices of the era. Earlier anecdotal references exist, including ancient accounts by Flavius Josephus of stampede-like crushes during Jerusalem festivals under Roman oversight, where structural exits and crowd momentum contributed to mass casualties, though numbers like 20,000-30,000 likely reflect hyperbolic historiography rather than empirical counts.

Modern and Recent Examples (1980–2025)

One of the earliest major crowd crushes in this period occurred on April 15, 1989, at Hillsborough Stadium in Sheffield, England, during an FA Cup semi-final soccer match between Liverpool and Nottingham Forest, where 97 Liverpool fans died from compression asphyxia due to severe overcrowding in perimeter fencing pens after police opened gates allowing too many supporters into already packed areas. An additional 766 people were injured, with the incident attributed to failures in crowd management and stadium design flaws that trapped fans against barriers. On July 2, 1990, during the Hajj pilgrimage in Mecca, Saudi Arabia, 1,426 pilgrims suffocated or were trampled in the Al-Ma'aisim pedestrian tunnel linking Mecca to Mina and Arafat, triggered by a combination of high crowd density, inadequate ventilation, and a sudden influx overwhelming the tunnel's capacity. The disaster highlighted vulnerabilities in linear infrastructure under mass movement, with air conditioning failure exacerbating heat and panic. In the 2010 Love Parade electronic music festival in Duisburg, Germany, on July 24, a crowd crush at the single narrow ramp entrance to the venue killed 21 attendees and injured over 500, as an estimated 1.4 million participants funneled into a confined space without sufficient exits or flow controls, leading to compressive forces that pinned victims against barriers. Investigations revealed permit violations and ignored capacity warnings as primary causal factors. The deadliest reported crowd crush of the era struck on September 24, 2015, during the Hajj stoning ritual at Mina near Mecca, Saudi Arabia, where Saudi officials recorded 769 deaths from a stampede and crush on the Jamarat Bridge, though compilations from affected nations by the Associated Press tallied at least 2,431 fatalities, reflecting underreporting discrepancies. Over 900 others were injured amid pilgrim flows exceeding infrastructure limits, with converging paths and poor signaling cited as triggers. On November 5, 2021, at the Astroworld Festival in Houston, Texas, a crowd surge during rapper Travis Scott's performance caused 10 deaths by compression asphyxia in a densely packed area near the stage, affecting an audience of about 50,000 where barriers failed to contain forward pressure from enthusiastic moshing. Hundreds more suffered injuries, with event planners having anticipated but not mitigated overcrowding risks. During Halloween celebrations on October 29, 2022, in Seoul's Itaewon district, South Korea, 159 people died in a crowd crush on a narrow, sloping alley accommodating around 100,000 revelers without adequate policing or density monitoring, as downhill pressure built from partying crowds funneled into bottlenecks. An additional 196 were injured, with official inquiries deeming it a preventable "man-made disaster" due to lapsed crowd control post-COVID restrictions. More recently, on July 2, 2024, at a religious gathering in Hathras district, Uttar Pradesh, India, honoring preacher Bhole Baba, a stampede killed 121 devotees—mostly women and children—after the event exceeded permitted attendance of 80,000, with chaotic exit rushes on uneven terrain and insufficient space causing compressive falls and trampling. Police reports attributed it to massive overcrowding and poor venue suitability, underscoring recurring issues in unregulated mass religious assemblies.

References

  1. https://pmc.ncbi.nlm.nih.gov/articles/PMC6308419/
  2. https://bpspsychub.onlinelibrary.wiley.com/doi/10.1111/bjso.12666
  3. https://arxiv.org/abs/0708.3339
  4. https://www.researchgate.net/publication/316286708_It%27s_a_crush_It%27s_a_collapse_It%27s_Wait_that%27s_No_Stampede
  5. https://www.sciencedirect.com/science/article/pii/S0925753523002345
  6. https://link.aps.org/doi/10.1103/PhysRevE.75.046109
  7. https://www.cambridge.org/core/journals/prehospital-and-disaster-medicine/article/survey-of-crowd-crush-disasters-and-countermeasures/99D8A45ED3C3F303B5E1CEFC454383BC
  8. https://www.workingwithcrowds.com/wp-content/uploads/2018/02/THE-CAUSES-AND-PREVENTION-OF-CROWD-DISASTERS-by-John-J.-Fruin-Ph.D.-P.E..pdf
  9. http://www.gkstill.com/ExpertWitness/CrowdDisasters.html
  10. https://www.theguardian.com/world/2022/nov/01/how-do-crowd-crushes-happen-stampede-myth-what-happened-in-the-seoul-itaewon-halloween-crush
  11. https://trueprepper.com/crowd-crush/
  12. https://riskfrontiers.com/insights/behaviour-and-mechanics-of-crowd-crush-disasters/
  13. https://www.theguardian.com/world/2015/oct/03/hajj-crush-how-crowd-disasters-happen-and-how-they-can-be-avoided
  14. https://www.crowdrisks.com/uploads/1/4/6/1/146195766/2000_still.pdf
  15. https://www.nfpa.org/news-blogs-and-articles/nfpa-journal/2020/07/01/crowd-control/too-close-for-comfort
  16. https://theconversation.com/ten-tips-for-surviving-a-crowd-crush-112169
  17. https://www.massey.ac.nz/about/news/how-data-based-tools-can-prevent-crowd-crushes/
  18. https://pubmed.ncbi.nlm.nih.gov/28372525/
  19. https://www.researchgate.net/publication/315789935_Acute_forces_required_for_fatal_compression_asphyxia_A_biomechanical_model_and_historical_comparisons
  20. https://pmc.ncbi.nlm.nih.gov/articles/PMC9995744/
  21. https://www.sciencedirect.com/science/article/pii/S2666449622000664
  22. https://www.sciencedirect.com/science/article/abs/pii/S2212420924003637
  23. https://www.nature.com/articles/s41586-024-08514-6
  24. https://epjdatascience.springeropen.com/articles/10.1140/epjds7
  25. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0306764
  26. https://www.jesip.org.uk/wp-content/uploads/2022/03/Hillsborough-Stadium-Disaster-final-report.pdf
  27. https://assets.publishing.service.gov.uk/media/5a7b2883e5274a34770e9d1d/understanding_crowd_behaviour-supporting-evidence.pdf
  28. https://pmc.ncbi.nlm.nih.gov/articles/PMC3426156/
  29. https://www.webmd.com/a-to-z-guides/stampede-what-to-know
  30. https://www.get-licensed.co.uk/safety-hub/crowd-safety-tips/
  31. https://www.sciencedirect.com/science/article/abs/pii/S1569190X24000923
  32. https://www.researchgate.net/publication/390185938_Analysis_of_the_Efficiency_of_Slope_Evacuation_for_Dense_Crowds_in_Urban_Street-Type_Public_Spaces
  33. https://pmc.ncbi.nlm.nih.gov/articles/PMC12011762/
  34. https://www.cambridge.org/core/journals/prehospital-and-disaster-medicine/article/human-stampedes-an-updated-review-of-current-literature/CB36E536DB49060879F3CB498A469498
  35. http://www.gkstill.com/CV/Modelling/CrowdCollapse.html
  36. https://www.npr.org/2021/11/10/1054428087/after-astroworld-a-crowd-scientist-explains-the-deadly-dynamics-of-crowd-surges
  37. https://scroll.in/article/852464/its-not-panic-that-kills-a-basic-guide-to-detecting-signs-of-a-stampede-and-escaping-it
  38. https://www.sciencedirect.com/science/article/abs/pii/S2212420917302212
  39. https://pmc.ncbi.nlm.nih.gov/articles/PMC11456318/
  40. https://www.workingwithcrowds.com/compression-asphyxia/
  41. https://www.sciencedirect.com/science/article/pii/S2212420924004151
  42. https://www.undrr.org/understanding-disaster-risk/terminology/hips/so0302
  43. https://www.thehindu.com/sci-tech/health/all-you-need-to-know-about-compressive-asphyxia/article70120638.ece
  44. https://www.pbs.org/newshour/health/how-and-why-do-crowd-surges-turn-deadly
  45. https://pubmed.ncbi.nlm.nih.gov/39371751/
  46. https://pubmed.ncbi.nlm.nih.gov/9810257/
  47. https://pmc.ncbi.nlm.nih.gov/articles/PMC4919827/
  48. https://www.ovid.com/journals/dmphp/fulltext/10.1097/dmp.0b013e3181c5b494~human-stampedes-a-systematic-review-of-historical-and
  49. https://www.researchgate.net/publication/7581432_The_pathological_features_and_circumstances_of_death_of_lethal_crushtraumatic_asphyxia_in_adults-A_25-year_study
  50. https://pubmed.ncbi.nlm.nih.gov/20081418/
  51. https://pmc.ncbi.nlm.nih.gov/articles/PMC11650299/
  52. https://www.researchgate.net/publication/378222562_Global_Compilation_and_Analysis_of_Human_Crush_Fatalities_1900-_2017
  53. https://www.sciencedirect.com/science/article/pii/S0925753523001169
  54. https://sgsa.org.uk/physical-factors/calculating-safe-capacity/
  55. http://www.gkstill.com/Support/crowd-density/CrowdDensity-1.html
  56. https://www.nfpa.org/news-blogs-and-articles/blogs/2016/12/07/life-safety-basics-in-assembly-occupancies
  57. https://up.codes/s/minimum-corridor-width
  58. https://www.structuremag.org/article/balustrade-design-loads-failures-fatalities-research-global-design-practices/
  59. https://www.researchgate.net/publication/333716797_Analysis_of_structures_subjected_to_crowd_loads
  60. https://www.nfpa.org/news-blogs-and-articles/blogs/2022/11/01/strategies-for-crowd-management-safety
  61. https://law.tmchippewa.com/us/nsn/tmchippewa/council/code/50.07.060
  62. https://deltascientific.com/2024/04/23/the-role-of-barriers-in-preventing-crowd-surges/
  63. https://files.ifea.com/RiskMgt/2019-08-13ERMSBriefDesigninginCrowdSafety.pdf
  64. https://www.oasys-software.com/news/dynamic-crowd-modelling-better-stadium-design/
  65. https://ww3.rics.org/uk/en/modus/built-environment/resilient-infrastructure/crowd-control--making-large-venues-safer.html
  66. https://www.hse.gov.uk/event-safety/crowd-management-control-inside.htm
  67. https://www.osha.gov/sites/default/files/publications/Crowd_Control.pdf
  68. https://www.researchgate.net/publication/373765037_Ranking_and_Analysis_the_Strategies_of_Crowd_Management_to_Reduce_the_Risks_of_Crushes_and_Stampedes_in_Crowded_Environments_and_Ensure_the_Safety_of_Passengers
  69. https://www.fielddrive.com/blog/navigating-legal-and-regulatory-requirements-in-event-security-planning
  70. https://www.journalhistoire.fr/quand-lyon-donna-sa-part-a-dieu/
  71. https://en.anecdotrip.com/la-guillotiere-a-croco-and-a-ripple-through-the-crowd-by-vinaigrette
  72. https://www.britannica.com/event/Hillsborough-disaster
  73. https://www.policeconduct.gov.uk/our-work/investigations/hillsborough
  74. https://www.bbc.com/news/uk-england-merseyside-47697569
  75. https://www.history.com/this-day-in-history/july-2/pilgrim-stampede-kills-1400
  76. https://www.vox.com/2015/9/24/9392929/hajj-stampede
  77. https://www.sciencedirect.com/science/article/pii/S0925753523001716
  78. https://www.bbc.com/news/world-europe-26152045
  79. https://apnews.com/general-news-3a42a7733a8b476889bb4b7b3be3560e
  80. https://www.reuters.com/article/world/india/death-toll-in-saudi-haj-disaster-at-least-2070-reuters-tally-idUSKCN0SN2FT/
  81. https://www.bbc.com/news/world-asia-india-34483500
  82. https://www.bbc.com/news/entertainment-arts-68618648
  83. https://abc13.com/post/astroworld-timeline-what-happened-at-concert-crowd-crush/13441821/
  84. https://houstonlanding.org/astroworld-planners-foresaw-crowding-before-deadly-festival-no-way-we-are-going-to-fit-50k/
  85. https://www.theguardian.com/world/2022/oct/31/how-did-the-seoul-itaewon-halloween-crowd-crush-happen-unfolded-a-visual-guide
  86. https://www.nbcnews.com/news/world/south-korean-police-officers-jailed-man-made-halloween-crowd-crush-rcna173387
  87. https://www.sciencedirect.com/science/article/pii/S092575352400331X
  88. https://www.aljazeera.com/news/2024/7/3/what-caused-the-deadly-crowd-crush-in-hathras-india
  89. https://www.bbc.com/news/articles/cv2g7wgj83no
  90. https://www.reuters.com/world/india/massive-overcrowding-led-stampede-indias-hathras-document-says-2024-07-03/
Add your contribution
Related Hubs
User Avatar
No comments yet.