Over the past few years, bollards have become newsworthy as a strategy against terrorist attacks. In the wake of tragedies, many of the questions asked about these little traffic posts are about their inherent stopping power. Yet there are many types of bollards, and each has a different application. Even with impact-resistant bollards, there are a range of intended behaviors. Physics, material science, and engineering all play into how much impact a bollard can take and disperse. What is often a surprise is that a bollard’s impact resistance has more to do with installation than with the bollard itself.
There are two main ways of installing a bollard. Many bollards are bolted down: low-impact, flexible, removable, and decorative bollards can be affixed on a plate that is attached to the ground with concrete anchors, or bolts. These installations mostly provide traffic guidance rather than security. Some provide minor impact-protection for slow moving small machines like floor-cleaners and fork-lifts.
The other group of bollards are embedded. Not all embedded bollards give substantial impact protection. Some, like retractable bollards, have an embedded mount; when raised, these are essentially surface bollards, with no substantial footing in the substrate.
Bollards with deeply embedded footings provide impact advantage. Their resistance is based on the physics of the portion of the bollard that lies beneath the ground.
Crash-rated embedded bollards have specially engineered footings sunk into the substrate. These bollards have been tested to have stopping power against vehicle incursion. When installed property their test results are given as their crash-rating.
More common than crash-rated bollards are security bollards. These also are dug into the ground and provide impact protection. However, they do not have an engineered footing. Since substrates and installations differ from site to site, security bollards have variable stopping power, which is why they are not rated.
A simple machine
Common security bollards are made of steel pipes filled with concrete. When such a bollard is sunk into the ground it creates one of physics’ most simple machines: the class one lever. This lever changes the direction of the force by pivoting around a fulcrum. When a bollard is sunk into the substrate, the ground becomes the fulcrum. The bollard, impacted, will begin to pivot around the lip of ground. The buried part of the bollard pushes back against the concrete and earth surrounding it, attempting to “lift” all the earth in the opposite direction. If dug deep enough, anchored in stable substrate, it will be unable to shift.
The nature of the substrate and the depth of the bollard both contribute to how well the bollard can receive and distribute impact forces. The effectiveness of the lever is constrained, however, by the strength of the materials.
A lever can shear or fail when stressed. Archimedes said of levers, “Give me a place to stand and I shall move the world”—but this good place to stand would not be enough to do the job. Archimedes would also need a lever made of a material strong enough to stay unbending under its own weight, and then additionally the mass of the Earth. On a much smaller scale, it’s the same with bollards: there is a point at which creating a longer bollard sunk more deeply into the ground will not help the bollard’s stopping power because the materials will not withstand the impact force.
However, many security bollards are made of steel, and steel is a very strong material. In most well-engineered bollard installations, it’s not the bollard that will fail. Rather, it’s the fulcrum. The bollard pushes against the ground opposite the impact, and that ground must also withstand the force of the impact. For bollards, that’s the concrete at or near the surface of the installation.
Steel vs. concrete
When a bollard is hit, it transmits the impact force to its fulcrum. For a solitary bollard in unreinforced concrete, this is the concrete edge of the installation.
It is almost universally the concrete edge, and not the steel of the bollard, that fails first.
This failure may be predictable, but it’s not possible to perfectly describe how the bollard and footing will behave. An engineer can evaluate only likely ranges of behavior, given different vehicles and speeds. Engineering analysis is complex in these situations. This is why engineers design for a large safety factor during specification.
Materials are tested in a laboratory to find stress-strain curves. These curves show the relationship between external stresses placed on a material and the strain or deformation within it. At the end of the stress-strain curve is the rupture point of the material, where it fails. These stress-strain curves are well known for steel and concrete.
Still, the real world is more complex than just comparing the two curves and predicting what the crash will look like. An impact between a car and a bollard is a moving system with feedback loops and changing variables. The behavior of the substrate determines how far a bollard can move upon impact. It may only move a few millimeters. Even if that is the case, the metal may travel far along its stress-strain curve, to areas of elastic and plastic deformation. As the crash continues the angle of force being applied to the bollard changes and the bollard moves and the vehicle decelerates. The steel of the bollard may move into its “non-linear property region,” where slight changes in stress mean different behavior in the steel. It becomes much harder to predict overall behavior.
The natural and forced frequencies of the part being hit are also important. Natural frequency is the frequency at which a system oscillates in the absence of any driving or damping force, like the resonation of a tuning fork. Such a natural frequency exists over the bollard amid driving and damping forces: the impact vector of the vehicle, resonance of the incoming material, and damping of the overall environment come into play.
In order to model this complicated real-world scenario, engineers can use computer-modeled finite element analysis (FEA) to estimate von Mises stresses in a system.
Von Mises stress modeling
Von Mises stress values are used to predict the failure point of material. In FEA, known stress-strain information and various force vectors are combined into a mathematical mesh laid over a simulated structure. This computer modeling can create an overall view of the system and make some generalized predictions about risk levels and failure points throughout.
In the FEA simulations above, we see a bollard embedded in concrete, flexing at the time of impact. The movement of the bollard is greatly exaggerated for visual effect; actual movement is less than 2mm. The color represents von Mises stress, with blue being the unstressed state, and red capturing point of failure. On the bollard, we can see impact stress on the bollard both on the side that is hit (Fig. 2) and on the side that’s flexing (Fig. 1). The lever action of the bollard also evidences stress beneath the substrate, where the buried end of the bollard is moving in an arc (Fig. 2).
The greatest points of stress, however, are those in the concrete at ground level. The red of the FEA modelling shows failure, not just on the far side of the bollard, but wrapping 2/3 of the way around toward the oncoming vehicle.
This simulated crash demonstrates the overall trend, that a heavy, deep-set steel pipe will withstand impact better than the concrete around it. The depth of the bollard, and the material of the substrate also have their parts to play. In general, however, to create an impenetrable bollard, reinforcing the footing around a bollard is necessary for greater crash resistance. In all cases, a site engineer must make a model and a prediction about how the bollard will do in its environment.
When failure is intended
Some installations require a high level of protection. Buildings that could be the targets of terrorism may need a deeply reinforced network of bollards that are tied together in a subterranean structure that guarantees stopping power.
However, impenetrable barriers can have a downside. Although cars and trucks are built with crumple zones and other safety features, bollards with superior stopping power may be the last object standing in a collision. The incoming vehicle—and its driver—might take the damage.
In some places at risk for being targeted, this may be necessary. In other places, like a sharp bend in the road that may be missed by motorists, or at the entrance to a store where pedal error is common, a little give might be desirable. Though the bollard may bend or “fail” while stopping the car, it may be working perfectly for the situation. Replacing a bent bollard or two and having the driver walk away without injury, in many cases, is the best outcome.
Communicating intention to an engineer
Crash-rated bollards are engineered so that neither bollard nor footing will fail when the system is hit by a vehicle at speed. These ratings are not always needed nor wanted in situations where a bent bollard does its job of preventing a small accident from becoming a big issue. However, because every site is different, only an engineer with knowledge of expected forces, content of substrate, depth of bollard, and material used in installation will be able to predict what is likely to happen in the event of a crash. Solid steel bollards have a lot of strength and protection to offer. It is the installation that determines how that strength will behave in a crash scenario.