On Safety&Security Supplement of March/April 2013 you can read an article dedicated to S.M.A. crash cushions. The title of the article is: Barrier Safety First.
SMA (Safety Modular Absorber) crash cushion: the first redirective crash cushion designed and made in Italy

Introduction

Industry A.M.S. srl has designed and patented a family of redirective crash cushions able to stop a vehicle from 50 Km/h up to 110Km/h. Recently it also get the CE certification for the parallel version, whereas for the V-shaped version the experimental tests are now running.

A robust design approach is utilized indeed until now the SMA (Safety Modular Absorber) crash cushions have successfully passed twelve experimental tests, each one at the fist attempt. This witnesses the reliability of the project. In figure 1 the family of crash cushion is displayed for the levels of velocity of 50, 80, 100 and 110 Km/h.

The design of the SMA crash cushion was oriented to match the following targets:

• High security: the head injury criteria (HIC) is monitored even if it is not expressly required by the European normative; the HIC level never overcomes 200 and this allows that the maximum abbreviated injury scale is always less than 2; this means that the driver does not undergo serious injury like fracture of the sternum.
• Reusability: the components of the crash cushion not devoted to the energy absorption are designed in order to preserve their original shape after an impact under the condition of the 1317. This means that after an impact only the crash absorbers should be replaced. This is of particular interest for the road manager in order to reduce the maintenance costs.
• Length: the high efficiency of the crash absorbers allows to minimize the crash cushion length. The reduced length allows using crash cushions in places where before it was not possible: a typical example is the by-pass in the tunnels. Furthermore our recent study clearly shows that the probability to impact a crash cushion increases as the crash cushion length increases, simply because the crash cushion itself behaves like an obstacle on the roads. A 3 meters crash cushion is compared with a  meters crash cushion with respect to the effect on the probability to impact the crash cushion itself. The study regards an installation in correspondence of a gate of the highway SS 162 NC Asse Mediano: the vehicle flux is assumed to be 3000 vehicle per day, the speed limit is fixed to 80 Km/h. In these conditions the probability to get an impact is 14 impacts per year for the 5 meters crash cushion and it is 14 impacts per year for the 3 meters crash cushion. This means that the probability to impact a crash cushion is simply proportional to its length: shorter crash cushion is a safer crash cushion.
• Easy installation: the crash cushion is constructed in order to be delivered already assembled in each part; two installation options are available: the first one by using connecting screws (the nuts are embedded in the concrete basement) and the second one through chemical anchors. The first option avoid obstruction on the floor and it is particularly suitable to resist to repeated impacts. It is at study the possibility to fix the system directly to the asphalt with chemical anchors.
• No maintenance: the use of material with an high durability is important to reduce the maintenance problems; the crash cushion is made of galvanized steel in each part consequently it does not requires any adjustments if there are no impacts and matches the fire class zero. The latter aspect allow to use the crash cushions in the tunnels.
• Low cost: the product cost is mainly affected by the material cost and the process cost; Using carbon steel and an automatized process allows contractin the product cost at a minimum in the category of crash cushions. The low direct cost is also accompanied by low indirect cost related to the installation, reusability, maintenance, low probability to get an impact (due to the reduced length)

General technical aspects

The problems involved in the development of a crash cushion can be schematized as the problem to stop a certain mass moving at an initial velocity V0 in a certain space S.

The European normative EN 1317 defines a limit for the maximum deceleration of the vehicle during the impact of a crash cushion through the definition of the ASI parameter.

During the impact the force required to deform the crash cushion affects cushion the level of deceleration of the vehicle, simply because the inertial force Fa= m a of the vehicle (where m is the mass of the vehicle and a the deceleration) equals at each time the force, F, required to deform the crash cushion. Generally crash cushions contain energy absorbers that transform the kinetic energy of the vehicle in to internal energy of the absorbers. It is useful to define the efficiency, η, of the energy absorbers in terms of force as follows.

1. where F(x) is the force required to deform the energy absorber, x is the current deformation of the energy absorber and S is the maximum deformation that undergoes the absorber, F_m is the average force and F_max is maximum force acting on the vehicle. If the absorber works during an impact with a vehicle, it is clear that the efficiency of the absorber equals the ratio of the mean deceleration during the impact and the maximum deceleration during the impact. Referring to a frontal impact the EN 1317 gives us a limit for a_max: a_max<16.8 g = 12g ASI_max, where ASI_max=1.4. According to this approach the minimum length of crash cushion can be easy predicted, once the desired ASI and the efficiency of the energy absorber are known:
2. Equation (2) shows clearly that, independently to the level of velocity (V₀ = 50, 80, 100 e 110 Km/h) and fixed the level of the severity of the impact (ASI, in any case < 1.4), the length of the crash cushion (S) is manly affected by the efficiency of the crash absorbers (h): for the same level of velocity a shorter crash cushion corresponds to an higher efficiency.

The AMS crash absorber: a metallic honeycomb

Industry A.M.S. srl has developed and patented a crash absorbers with high efficiency as defined according to equation (1). The crash absorber consists of a metallic honeycomb made using a stamping and welding technology from metal sheets. The process of fabrication of the steel honeycomb panels is completely automatic and allows using these panels in all the situations where a low cost is required.

The force versus displacement curve reported in Figure 2 is like to a step function, this means that the steel honeycomb behaves almost as a perfect energy absorber. This aspect is of a particular relevance looking at the biomechanical parameters that allow evaluating the injury of the driver during a impact. As long as the deceleration diagram does not present peek the biomechanical parameters, like the head injury criteria (HIC), are constrained to values such that the driver experiences no severe injury. Furtermore, the efficiency of the energy absorbed is calculated to be about 0.92 (92%). This high efficiency allows the construction of crash cushion with a very small length according to equation (2). The minimum length of a crash cushion can be easy calculated from equation (2) assuming an efficiency of 100%. This means that a perfect and ideal crash cushion could be shorter than the SMA one of only 8%.

SMA crash cushion family

Crash cushions for impact velocities of 50 km/h, 80 km/h, 100 km/h and 110 km/h have been designed using as the steel honeycomb energy absorber. Due to the high efficiency of the utilized energy absorber we were able to reduce the size of the crash cushions to 2.0 m, 3.3m, 5.1m and 6.5m for the 50 km/h, 80 km/h, 100 km/h and 110 km/h, respectively. It shall be pointed out that the above mentioned lengths coincide with then operative lengths and consequently the SMA crash cushion can be placed near the obstacle to protect: no additional space is required to install the crash cushion.

In the figures 4a-4c some images get from the homologation test TC.1.3.110 in three different instants are reported. This crash test is the more severe one in terms of energy, mass and velocity of the entire family. The deformation of the crash cushion occurs progressively and tardily without abrupt variation of the motion of the vehicle. The figure 4d shows that, after the impact, the cockpit is almost undeformed so that the vehicle door can be easy opened by the driver. This feature together with the results in terms of biomechanical parameters allow concluding that the driver, after an impact at 110Km/h with the SMA crash cushion, will be able to leave the car with his legs without waiting the help of the relief people.

In the figure 5 the ASI diagram versus time is reported for both the experimental and the numerical test. It should be noted that the numerical calculation are able to predict pretty well the experimental behaviour of the crash cushion, indeed the two curves reported in figure 5 are almost superposed. Furthermore the ASI curve is very flat without any peak. This feature excludes the possibility to get peaks in the deceleration diagram that could jeopardize the safety of the driver.

The results in terms of deceleration of the centre of mass of the vehicle obtained during the twelve crash test carried out at the CSI spa proving ground have been used to calculate the head injury criteria (HIC). The HIC has been obtained through a numerical simulation of the sled test with a hybrid III dummy. HIC is a biomechanical parameter based on the acceleration diagram of the centre of gravity of the dummy. It has been showed that the HIC is well correlated to the head injury of the driver. The results of these calculations in terms of HIC versus ASI are reported in figure 6 together with the diagram of the MAIS (Maximum Abbreviated Injury Scale) versus the HIC. Figure 6 clearly show that the calculated HIC is always lower than 200 irrespective of the crash test type and that the MAIS is at most equal to 2 (with a probability of 20%). This last feature excludes severe injury of the driver.