Checked against current official sources: 5 September 2026.
A road splits.
One lane goes left.
One lane goes right.
Between them sits a hard point.
Maybe it is the nose of a concrete divider.
Maybe it is a barrier terminal.
Maybe it is another roadside object that cannot simply disappear.
A driver makes a mistake.
The vehicle is now heading directly toward the hard point.
A crash cushion works because the collision is going to spend energy somewhere, and the road can choose a sacrificial device to spend some of it before the vehicle reaches the harder object behind.
Quick Read
Singapore works partly because roadside design does not assume every driver will stay perfectly inside the intended path.
LTA’s current Road Safety guidance says a crash cushion is designed to absorb the impact of a collision, reducing injury to passengers and minimising damage to the vehicle. LTA places crash cushions in front of diverging roads and along expressways and major arterial roads.
The deeper mechanism is:
vehicle leaves intended path → vehicle reaches crash cushion before the hard roadside hazard → cushion deforms, collapses, strokes or otherwise dissipates energy according to its approved design → stopping distance and stopping time are increased compared with an abrupt rigid impact → peak forces on vehicle and occupants can be reduced → vehicle loses speed before reaching the protected hard point → collision severity is lower than it would have been against the rigid obstacle directly.
This article does not claim that crash cushions make collisions harmless, that every impact occurs within the device’s design envelope, or that every crash cushion uses one identical energy-absorbing mechanism. It isolates one overlooked mechanism: when a dangerous hard point cannot be removed, it can sometimes be preceded by a structure intentionally designed to fail more gently than the object it protects.
Wait, What? The Cushion Is Designed to Be Damaged?
Often, yes.
That is not a defect.
It is the mechanism.
A perfectly rigid object barely deforms.
The vehicle must then lose its momentum over a very short distance and time.
A crash cushion allows controlled deformation or displacement.
That deformation consumes energy and lengthens the deceleration event.
the device survives by doing its job; the job may require the device not to survive unchanged.
The Physics Begins with Kinetic Energy
A moving vehicle carries kinetic energy.
For a simplified model:
E = ½mv².
Mass matters.
Speed matters even more strongly because speed is squared.
A faster vehicle therefore brings disproportionately more energy to the crash cushion.
The device has to redirect that energy into deformation, friction, material crushing, vehicle deformation and residual motion rather than letting the protected hard point absorb the impact directly.
Stopping Distance Is a Safety Resource
If a vehicle must lose the same speed, doing so over a longer distance usually reduces average deceleration compared with stopping almost instantly.
The crash cushion therefore purchases safety with distance.
Its physical length is not wasted roadside space.
It is the distance over which collision energy can be managed.
when impact cannot be avoided, an extra metre of controlled deceleration can be more valuable than an extra metre of rigid structure.
The Crash Cushion and The Road Hump Own Different Moments
The Road Hump article owns speed reduction before a collision.
Its geometry persuades or forces drivers to approach more slowly.
The Crash Cushion owns energy management after the vehicle has already departed from the intended path.
Road hump:
reduce risky speed before the incident.
Crash cushion:
reduce collision severity after avoidance has failed.
Prevention and consequence control remain distinct layers.
The Crash Cushion and The Safety Bollard Own Different Receivers
The Safety Bollard article owns a protective line between errant vehicles and waiting commuters at bus stops.
The Crash Cushion primarily protects vehicle occupants from a severe direct impact with a fixed roadside hard point.
Safety bollard:
keep the vehicle from reaching people.
Crash cushion:
make the vehicle’s unavoidable contact with roadside infrastructure less abrupt.
Both use sacrificial energy management.
The protected receiver and geometry differ.
Why Diverging Roads Create Hard Points
At a road split, two traffic streams separate.
The area between them may contain a gore point, divider, barrier terminal or other fixed element.
That object cannot be moved indefinitely away because the road geometry itself needs separation.
A driver who commits too late to one branch can head directly toward the nose.
Placing a crash cushion before the hard point transforms the first object the vehicle encounters.
Instead of:
vehicle → rigid divider.
The sequence becomes:
vehicle → energy absorber → harder object behind.
The Device Is a Designed Failure Sequence
Good crash cushions do not merely “break.”
The approved design controls how the system deforms under expected impact conditions.
Depending on the system, energy absorption can involve:
- crushable elements;
- telescoping sections;
- friction;
- hydraulic or mechanical dissipation;
- material yielding;
- or combinations of these mechanisms.
The important point is not one proprietary mechanism.
It is sequencing:
soften first contact → absorb energy progressively → protect the harder terminal behind.
Alignment Matters Because a Cushion Protects a Trajectory
A crash cushion is useful only if the errant vehicle actually encounters it before the hazard.
That makes placement a trajectory problem.
The engineer asks:
- Where could a vehicle leave the lane?
- At what angle could it approach?
- What object lies behind?
- Can a vehicle bypass the cushion and strike the hard point anyway?
- Does the road geometry guide an errant path into the intended face of the device?
The strongest energy absorber in the wrong place is weaker than a correctly placed device designed for the likely path.
Head-On and Angled Impacts Are Different Problems
A vehicle striking the centre of a cushion straight on loads the system differently from a vehicle clipping it at an angle.
Different approved crash-cushion systems can be designed to respond differently to redirection, gating and energy absorption.
This article does not reduce all devices to one collision model.
The broader lesson is:
impact protection must be designed for a family of plausible trajectories, not a single perfect crash drawn through the centreline.
Occupant Injury Depends on Deceleration, Not Just Vehicle Damage
A dramatic-looking crushed front end can represent energy being absorbed by vehicle and roadside structures.
A vehicle that remains visually intact after striking an immovable object is not necessarily safer if the occupants experienced extreme deceleration.
Roadside safety therefore cares about the forces transmitted to people, not merely the amount of visible infrastructure damage.
LTA’s own description focuses on reducing injury to passengers as well as vehicle damage.
Vehicle Restraints and Crash Cushions Work Together
The roadside system does not act alone.
Inside the vehicle, seat belts, airbags and vehicle crumple structures manage occupant motion.
Outside the vehicle, the crash cushion manages the vehicle’s interaction with the hard roadside point.
Layered protection means:
road absorbs some energy → vehicle structure absorbs some energy → restraints manage the occupant’s remaining motion.
Safety emerges from the chain, not one miracle component.
Visibility Still Matters Before Impact
Crash cushions are consequence-control devices.
The best outcome is still no collision.
That means chevrons, markings, lighting, road geometry and advance signs remain important.
The crash cushion should be visible enough that the driver can avoid it when recovery remains possible.
the safest energy absorber is the one that never has to absorb energy because the driver recovered in time.
A Previous Impact Changes the Next Impact
A crash cushion that has already been struck may no longer have its original energy-absorbing capacity.
Crushable elements may be spent.
Frames may be distorted.
Anchors may have moved.
Reflective markings may be damaged.
One successful crash can therefore consume readiness for the next one.
sacrificial safety systems need a fast return from used state to ready state.
Inspection Is About Hidden Deformation Too
Damage is not always dramatic.
A low-speed strike can shift alignment.
A fastener can loosen.
Corrosion can weaken a component.
Road resurfacing can alter the surrounding height or approach geometry.
The device can still look like a crash cushion while its next-impact performance has changed.
The Hard Point Behind the Cushion Still Matters
The cushion does not remove the barrier terminal, bridge pier or divider nose behind it.
It manages the path into that object.
If the protected object changes, the crash-cushion requirement may change too.
Road modifications therefore need to consider the complete roadside system, not treat the cushion as independent furniture.
The Bottleneck Is Energy per Available Stroke
A cushion has finite physical length and finite deformation capacity.
The incoming vehicle has mass and speed.
The core problem is:
can enough of the vehicle’s incoming kinetic energy be dissipated over the available device stroke without transmitting unacceptable forces or allowing the vehicle to reach the hard point with too much residual energy?
This is the physical bottleneck hidden inside a simple roadside object.
Receiver: The Vehicle Occupant
The primary receiver is inside the errant vehicle.
That person may have:
- misjudged the split;
- fallen asleep;
- lost control;
- experienced a medical event;
- been struck by another vehicle;
- or reacted too late.
The road infrastructure does not ask whether the mistake was reasonable before trying to reduce the consequences.
This is failure-tolerant design.
Receiver: The Next Driver
After an impact, damaged crash-cushion components, a disabled vehicle or debris can create a secondary hazard.
Incident clearance and repair therefore protect the next road user as well as restore the original device.
A safety system has to return the road from:
collision state → protected recovery state → stable ready state.
Competing Explanation: Why Not Remove Every Fixed Roadside Hazard?
Where a hard point can be removed, relocated or made safely traversable, that can be preferable.
But some objects are necessary:
- barrier systems;
- bridge structures;
- road-dividing geometry;
- sign supports;
- and other infrastructure constrained by the road layout.
The crash cushion exists when the hazard cannot be eliminated adequately and an energy-absorbing treatment is appropriate.
Competing Explanation: Why Not Use a Concrete Block?
A concrete block can stop a vehicle.
Stopping is not the only objective.
The road-safety question is:
how does the vehicle stop, over what distance, with what deceleration and with what consequences for occupants?
A rigid block solves containment and can worsen occupant deceleration.
A crash cushion is designed to manage the stopping process.
Model Limit: Not Every Vehicle Fits the Same Impact Envelope
Cars differ in mass, height, stiffness and geometry.
Motorcycles behave differently.
Heavy vehicles carry much greater momentum and energy.
Impact angle and speed vary.
Approved roadside hardware is therefore selected against standards and expected operating conditions, not a fantasy of one universal vehicle.
Model Limit: A Cushion Cannot Fix Bad Upstream Geometry
If a road split repeatedly surprises drivers, the solution may also require:
- better advance signs;
- clearer lane markings;
- speed management;
- lighting;
- or geometric redesign.
The crash cushion treats the consequence path.
It should not become an excuse to ignore a preventable upstream error pattern.
What Breaks First?
- The device is misaligned with the likely impact path.
- The approaching vehicle bypasses the cushion and strikes the protected object.
- A previous impact has consumed energy-absorbing capacity.
- Anchors or structural elements are damaged.
- Road resurfacing changes the device’s effective geometry.
- Reflective or warning markings become difficult to see.
- The incoming vehicle exceeds the expected impact envelope.
- Damage is not detected and the device remains in service in a degraded state.
The useful Wintour House audit question is:
if a vehicle left the lane along the most plausible errant path now, would it encounter a ready, correctly aligned crash cushion before the hard point, and is enough deformation capacity available to reduce the severity of that impact?
Primary-School Lens: Soft Stop, Hard Stop
Use a drawing rather than a collision demonstration.
Draw a toy car heading toward a rigid wall.
Then draw a crushable box in front of the wall.
Ask which arrangement gives the car more distance to slow before reaching the hard wall.
The child learns that safe stopping can use deformation.
Secondary-School Lens: Why Speed Matters So Much
Compare the kinetic energy of the same vehicle at two speeds using E = ½mv².
Ask why doubling speed creates four times the kinetic energy in the simplified model.
Then ask what that means for a device with fixed deformation distance.
The road-safety problem becomes a direct consequence of the square law.
JC Lens: Work-Energy and Impulse
At JC level, the crash cushion can be analysed through work-energy and impulse.
The device does negative work on the vehicle as it deforms, reducing kinetic energy.
Increasing the time over which momentum changes can reduce average force for a given momentum change.
The engineering question becomes:
how should deformation force vary over device stroke so enough kinetic energy is absorbed while occupant deceleration, structural loads and residual vehicle motion remain within the accepted performance envelope?
Thought Experiment: Perfect Cushion Behind the Hazard
The device is excellent.
It is installed behind the concrete divider.
The vehicle hits the divider first.
Component quality succeeds.
Sequence fails.
Thought Experiment: Long Cushion, Very High Speed
The device offers substantial deformation distance.
The vehicle arrives with energy far beyond the expected envelope.
The cushion absorbs much of the impact and still bottoms out.
Safety device success is not the same as unlimited capacity.
Thought Experiment: Cushion Already Crushed
A vehicle hit the cushion yesterday.
The road reopened before the device was restored.
A second vehicle leaves the road today.
The first collision consumed the capacity the second collision expected.
Recovery is part of readiness.
Why Singapore Works Does Not Mean Crash Cushions Make Expressways Safe at Any Speed
Impact speed can exceed the design envelope.
Heavy vehicles can carry far more energy.
Vehicles can approach at unusual angles.
Devices can be damaged or misaligned.
Road geometry and driver behaviour remain primary safety layers.
The serious claim is narrower:
LTA uses crash cushions at diverging roads and along expressways and major arterial roads as impact-absorbing roadside devices intended to reduce passenger injury and vehicle damage by managing collision energy before an errant vehicle reaches a harder fixed hazard.
The crash cushion does not prevent the mistake.
It changes what the mistake meets first.
The Fifteen-Question Crash Cushion Test
- Hard point: What fixed hazard is being protected?
- Trajectory: What errant vehicle paths are plausible?
- Placement: Does the cushion intercept those paths first?
- Impact envelope: What speed, mass and angle range is the system designed around?
- Stroke: Is enough deformation distance available?
- Energy: Can the system dissipate the expected kinetic energy?
- Redirection: How does the approved device behave under angled impact?
- Anchorage: Are foundations and attachments intact?
- Visibility: Can drivers identify the road split and device early?
- Bypass: Can a vehicle miss the cushion and hit the hard point directly?
- Previous impact: Has the device already spent part of its capacity?
- Road works: Has resurfacing or reconstruction changed geometry?
- Inspection: Is hidden damage being detected?
- Recovery: How quickly is a struck cushion restored?
- World return: Do actual collision and near-miss patterns still match the hazard model used to place the device?
Frequently Asked Questions
What does LTA say a crash cushion does?
LTA says it is designed to absorb the impact of a collision, reducing injury to passengers and minimising damage to a vehicle.
Where are crash cushions used in Singapore?
LTA says they are placed in front of diverging roads and along expressways and major arterial roads.
Is a crash cushion the same as a safety bollard?
No. A bus-stop safety bollard primarily protects waiting commuters from an errant vehicle. A crash cushion primarily manages the vehicle’s impact with a fixed roadside hard point.
Why not simply use a rigid barrier?
A rigid barrier can stop or redirect vehicles but may create a severe direct-impact point at its terminal. A crash cushion is specifically used to manage impact energy at such exposed locations.
Does a crash cushion work more than once?
That depends on the approved system and severity of impact. Any struck device needs inspection, and sacrificial or damaged components may require replacement before full protection is restored.
What is the main student lesson?
Safety can depend on deciding what should deform first. When impact energy cannot be avoided, a controlled sacrificial path can protect people from a harsher uncontrolled one.
Sources and Further Reading
Final Thought: The Road Chooses What the Mistake Hits First
The driver should stay in the lane.
The signs should be clear.
The markings should guide.
The speed should be manageable.
And still the road prepares for the miss.
Not with another lecture.
With a structure willing to spend itself.
That is why Singapore works, in another quiet way:
the city understands that resilient design does not stop at telling people where not to go; it also asks what they should meet first if, one day, someone goes there anyway.