Low Speed Impact Crashes:

Engineering Analysis



Clifford R. Tyner, P.Eng

President, C.R. Tyner and Associates, Ltd.

and Stuart D. Smith, P.Eng

Associate Consultant, C.R. Tyner and Associates, Ltd.

Dartmouth, N.S.

September, 1998



A typical low-speed accident

Several cars are stopped at a traffic light. The light turns green and the first vehicle, driven by Mrs. Target, proceeds into the intersection, closely followed by Mr. Bullet in the second vehicle. A vehicle on the crossing road, attempting to make it through on the amber light, continues through the intersection on a red light and Mrs. Target brakes sharply to avoid it. Mr. Bullet, who is talking to his girlfriend on a cellular phone, is unable to stop and runs into the back bumper of the Target vehicle. There are minor scuffs on the bumper covers but no other visible damage to either vehicle. Afterwards, Mr. Bullet and Mrs. Target exchange information and continue on their way.

The next day Mrs. Target goes to her doctor with neck pains. He finds no visible symptoms and advises her to stay home from work for a couple of days. Subsequently, soon after the event, Mrs. Target complains of chronic neck pain and pursues a claim for long-term disability.

There are circumstances in which a serious injury may arise from a low speed accident, or an accident that causes minimal damage to the vehicles involved. However, with the public perception that injury claims will be compensated, a substantial number of injury or disability claims are arising from low-speed accidents, and not all of these claims may be legitimate. Our present challenge is to determine whether or not a particular accident was severe enough to give rise to a significant possibility of injury.

Four common collision types

By far the most common type of low-speed accident is a "rear-end" collision where the front of one vehicle runs into the back of the one ahead. Most often, it is the driver behind who is found to be at fault for failing to stop. Less frequently however, one vehicle may back into another. The human neck can withstand less vehicle acceleration (when the head is thrown back) than deceleration (when it is thrown forward), and consequently, so-called whiplash injuries are more likely to occur in the vehicle that is "rear-ended". In the discussion to follow we will deal with rear-end collisions unless otherwise noted.

Front-end or head-on collisions can occur at low speeds, but often these accidents occur at higher speeds; a common example of this is when a driver crosses the centre line of a highway and collides with an oncoming vehicle. Low-speed frontal collisions throw the occupants of both vehicles forward. The human neck can withstand stronger vehicle deceleration in a frontal impact than vehicle acceleration (when the head is forced backward in a rear end impact), and so a belted occupant can withstand a higher speed frontally than from a rear end collision. It is therefore less probable that there will be injuries in a low-speed frontal collision with negligible damage to the vehicles involved.

Lateral or "T-bone" collisions, tend to occur at intersections. This is where a vehicle enters an intersection and is run into by traffic travelling on the crossing road, or where a vehicle enters the intersection and runs into the side of another vehicle. Lateral collisions also commonly occur when a vehicle exits a driveway in front of oncoming traffic. The sides of most vehicles are much weaker than the front or rear. One of the most serious consequences of lateral collisions for occupants is due to the intrusion of damage on the side that is struck. Lateral collisions can occur at low or high speeds, but typically occur at intermediate speeds of urban traffic.

In sideswipe collisions, unlike all of the other types mentioned above, the vehicles come into sliding contact and do not come to a common velocity during the impact. For given vehicle speeds sideswipes produce a smaller change of velocity (D V) than do other types of impacts. The duration of a sideswipe event is longer than that of a direct impact, and a given D V results in smaller occupant acceleration. Unless there is intrusion into the passenger space a low-speed sideswipe is considered to be relatively benign to the occupants of the vehicles. However, a sideswipe may cause a loss of control, followed by a second collision and/or rollover that can have much more serious consequences.

"Plastic" and "elastic" collisions

In a "plastic" collision, both vehicles come to a common speed. If both vehicles are of equal weight and the "target" vehicle was stopped, the common speed is half the speed of the "bullet" vehicle. In high-speed collisions crushing the structure of the vehicles absorbs much of the energy. These collisions are mainly plastic, i.e., there is little rebound, and both vehicles often remain engaged during their post-impact movement from the impact area to their resting position.

In a purely "elastic" collision, on the other hand, none of the energy is absorbed. If two elastic objects of equal weight (e.g. billiard balls) collide, the first object stops and the struck object moves forward at the same speed at which it is struck. Here the term "elastic" means that the objects are springy and all of the impact energy is returned on rebound. It does not mean that the objects are stretchy.

A low-speed impact between vehicle bumpers is partly plastic and partly elastic. Within the limits of their design, bumper systems absorb the energy of the collision. Part of this energy is dissipated within the bumper system, and the remaining part is returned on rebound. The relative speed (between the vehicles) after a low speed impact is usually 30 – 70% of the relative speed before. The ratio of the relative speeds before and after is called the "coefficient of restitution". This coefficient tends to be larger for very low speed collisions than for collisions approaching the design limits of the bumper systems.

Figure 1, from Bailey et al., 1997, illustrates a typical low-speed rear-end collision of two cars of equal weight. In the top two rows the approach and impact speed of the "bullet" car (Vehicle 1) is V1 = 10 KPH and initially the "target" vehicle is stationary (V2 = 0). In the middle row the common velocity of both vehicles at the instant of maximum engagement is 5 KPH, half of V1. After impact and rebound the "target" car (Vehicle 2) has been accelerated to a speed V2 = 7 KPH, and the speed of the "bullet" Vehicle 1 has been reduced to 3 KPH. but less than V1. The severity of this accident to the occupants of Vehicle 2 is

D V2 = V2after – V2before = (7 – 0) = 7 KPH.

We assume that neither vehicle is accelerating or braking, and the average speed of the two vehicles remains the same (5 KPH) throughout the event. The relative speed of approach of the vehicles was initially V2 – V1 = 10 KPH. At the instant of maximum bumper engagement both vehicles reached a common speed (5 KPH) and their relative speed was 0 KPH. After the bumpers rebounded the relative speed was V2 – V1 = 3 – 7 = -4 KPH. The 4 KPH separation speed is represented as a negative approach speed. The restitution coefficient of the combination of the two bumper systems is the ratio of the separation and approach speeds, CR = 4/10 = 0.4. In this example the bumper systems acted in a way that was 40% "elastic" and 60% "plastic" with regard to approach speed.

Meaningful characterizations of collision severity

Determining collision severity from physical evidence

We need to estimate the severity, which is the CHANGE in speed (D V) of the vehicle in which there is an injury or claimed injury to the occupants. The way in which we make this estimate depends on the type of evidence available. Several methods are discussed below. It may be possible to estimate speed by more than one method and to check the consistency of the methods. If we can estimate the change of speed of one vehicle involved in a collision, we can usually calculate the change of speed of the other vehicle by momentum analysis.

Tire marks

If the brakes of a vehicle were locked and it left a tire mark as it came to a stop, we can estimate its speed at the start of the skid from the length of the tire mark and the friction coefficient of the road. A skid to a stop from 10 KPH leaves skid marks approximately 0.6 metres (2 feet) long. The length of the skid goes up with the square of the speed: skidding to a stop from speeds of 50 and 100 KPH the same vehicle would leave marks 15 and 60 metres long, respectively.

If the brakes of a moving (bullet) vehicle were applied and left tire marks, we can estimate the change of speed during braking. In order to estimate the speed before braking, the speed at impact must also be known.

Time or distance of travel of one of the vehicles involved

If one driver accurately recalls the details of the accident, we can have them drive in exactly the same way and measure the time and distance. For example, they may have released the parking brake and rolled forward down a slight incline in Neutral along a known maximum distance or for a known maximum time. Or they may have released the foot brake and rolled forward with an automatic transmission in Drive for a known maximum distance or time. This type of estimate will stand up in court only insofar as the accuracy of our driver’s recollection of his actions is accepted or is supported by other witnesses.

Hydraulic bumper struts

Older (1970’s and 1980’s) vehicles often had bumpers mounted on shock-absorbing hydraulic struts. When these struts are compressed and return to their original length, scratch marks are usually made on the "pistons" of the struts. The strut compression can be found by measuring the length of the scratch marks. This can provide the most accurate estimate of speed change during a low speed collision event. Photographs 1 and 2 show the front bumper struts of a 1985 Plymouth Reliant that has struck the car ahead on a right turn ramp. The impact was to the right corner of the front bumper, and only the right strut was compressed. The measured length of the scratch marks (in this case 3.8 centimetres) has then been compared with the strut compressions made from a series of controlled impact tests in which the front of a 1985 Plymouth Reliant struck the back of a stationary vehicle at progressively increasing speeds. These test data are available from a database compiled by MacInnis Engineering, Vancouver, B.C. In this example we estimated that the Reliant was travelling at 4 KPH, and that the impact imparted a speed (D V) of 3 KPH to the car ahead.

Bumper damage threshold

Newer cars usually have "foam core" or "lattice core" bumpers that absorb some of the impact and then return to their original shape without leaving any visible sign of deformation. Some of these newer bumper systems can withstand fixed barrier impacts at speeds in excess of 15 KPH without damage. These bumper survival levels are much higher than those for the older bumpers using hydraulic struts. Higher impacts cause deformation of mounting struts that must be replaced if they have been deformed. We can estimate an upper limit of the collision speed from a damage threshold established by series of controlled impact tests of one or both of the types of vehicles involved (again from the database of MacInnis Engineering). If one of the vehicles is lighter or has weaker bumpers, its lack of damage can establish an upper limit for the speeds of both vehicles.

Re-enactment of an accident

If there are no data for the types of vehicles involved, it may be necessary to obtain identical vehicles and re-enact the accident, or to have this done by a laboratory that specializes in conducting controlled crashes.

For example, in a rear-end collision between a large truck and a small older-model car that gave rise to a whiplash injury claim, the only visible damage was a small dent in the hatchback of the car that matched the bumper height of the truck. This dent did not resemble any available data from databases for bumper-to-bumper impacts. In this case we located the owner of an identical car (except of course for the dent). We brought the actual truck and the test car to a large, level vacant parking lot. We carefully aligned the car with the truck bumper, and then rolled the car very slowly to impact with the truck, timing its travel along a measured distance, and found that we had made a dent in the hatchback that was similar in size to the dent from the accident. We then rolled the car at a walking speed, and made a much larger dent.

We were able to report that the severity of the accident corresponded approximately to an impact at the lower test speed, and certainly at less than the higher (walking) test speed. This placed the change of speed (D V) well below the threshold at which any symptoms were reported by volunteer test subjects in controlled tests.

Of course we compensated the owner for the damage to his car. In this case the time and expense involved in re-enacting the accident was justified in relation to the size of the potential claim.

Speed estimate from frontal crush

Where the frontal impact is severe enough to shorten the front part of the vehicle, we can:

The NHTSA database is quite comprehensive because crash testing is required if the vehicle is to be sold in the USA. Most of the crashes are into a rigid barrier at speeds of 47 – 56 KPH. Sometimes we have to use tests from a "clone" of the vehicle involved (e.g. Plymouth Reliant and Dodge Aries) or from a different model year, for which the vehicle specification is nearly identical. If there are no NHTSA data for a specific vehicle, the uncertainty in the "generic" formula can be quite large, up to + 18 KPH.

In low-speed accidents, where there is minimal damage except perhaps to the bumper system, the frontal crush method usually cannot be applied.



Occupant movement in a rear-end collision

In a rear end collision the "target" vehicle undergoes a sudden change in forward speed (positive D V). Initially the occupant is not accelerated, and finds himself moving backward within the vehicle. Pressure from the back of the seat accelerates the occupant’s torso forward. The head continues to moves back and the neck bends back until the head makes contact with the headrest, or until the muscles and ligaments of the neck stop the relative motion between head and torso. If the limits of voluntary neck motion are significantly exceeded, injury and pain may result. With a properly positioned headrest, studies have shown that very high forward acceleration can be tolerated without injury.

By the time the occupant reaches his maximum backward travel into the seat cushion and headrest, the vehicle may now be decelerating. The deceleration and the rebound from the cushion propel the occupant forward until resisted by the seat belt (or by contact with the interior of the vehicle if unbelted, possibly causing injury).

Figure 2, from McConnell et al., 1993, illustrates the typical sequence of motion of an instrumented volunteer in a low-speed rear-end impact.

In newer vehicles, forward movement of the occupant may also be arrested by deployment of an airbag. However, seat belt use is still vitally important to keep the occupant in place while the airbag deploys. If the occupant moves too far forwards, up, down or sideways, misalignment with the airbag can result in injury when it deploys and or a failure to provide protection.

Human tolerance and collision severity

Once the impact accelerations and speed changes are determined, they then can be compared with known data from controlled collisions using volunteer test subjects. Szabo and Welcher (1998) recently presented an updated summary of instrumented low-speed collisions, using volunteer test subjects from 18 series of experiments (see Figure 3). For rear-end collisions there is a threshold severity of D V = 7 to 8 KPH (4.5 to 5 MPH) below which the volunteers rarely felt any symptoms at all, and only 0.4% reported symptoms that persisted more than 24 hours. At Delta V’s of 8 – 18 KPH approximately 10% of volunteers reported symptoms lasting longer than 24 hours. Typical symptoms included headache, neck pain or backache. Out of 391 exposures, "no significant injuries were reported by the researchers in spite of multiply exposed subjects (struck more than once over a one or two day period). In most cases no symptoms were reported. At most a week of mild neck pain has been reported" (Szabo and Welcher, 1998).

For frontal collisions, Szabo and Welcher report that "bullet vehicle drivers underwent D V’s of up to 11 MPH (18 KPH) without any complaints of pain".



Reconstructing a low-speed accident

In reconstruction of a low-speed accident, we generally follow three steps:

Basic information

In every accident it is important to know the year, make and model of vehicles involved. This information is normally contained in police reports, but occasionally there may be an error and it is valuable to have an independent record. The Vehicle Identification Numbers (V.I.N.) can be used to verify the weight, dimensions, power train and equipment of cars and light trucks. A builder’s plate, usually found on the driver’s door post gives the V.I.N., the Gross Vehicle Weight Rating (GVWR), the date of manufacture, and other valuable information. The V.I.N. is usually also found at the lower left corner of the windshield and on a label or stamping somewhere under the hood. It is also useful to note the odometer reading; mechanical inspection of older or high-mileage vehicles may reveal defects such as faulty brakes or worn-out steering components that can compromise the ability to avoid an accident. The expiry dates of the Motor Vehicle Inspection sticker and registration plate should be checked.

The name of the driver and the names and seating locations of passengers are important basic information. An estimate of the weight, height and age of each occupant is useful both for calculating vehicle dynamics and for estimating exposure to injury (e.g., was a very short driver seated close to an airbag, or did a tall driver have the seat positioned all the way back?).

The loading of the vehicle should be noted, particularly for trucks. Large trucks are often custom-made and the operator generally knows the empty-vehicle weight on each axle. The truck may have been weighed when it was loaded, or at a highway scale en route. If a car was heavily loaded with passengers and luggage, it may have exceeded the manufacturer’s gross vehicle weight rating (GVWR). This may cause detrimental effects on vehicle stability particularly under emergency steering and braking situations.

Vehicle damage: location and amount

The location of damage provides important information on the relative location and direction of travel of the vehicles when they came together or struck an object. Paint transfer can help to verify that the damage resulted from contact between specific vehicles and/or objects.

In low speed accidents the height of the top and bottom of the bumper(s) should be measured. It is also important to note the location of sharp objects such as license plate bolts that may have left marks on the other vehicle.

Photographing vehicle damage

We usually start an accident reconstruction by taking measurements at the scene and by inspecting the vehicles involved. If the vehicles are available they should be inspected without delay. At a later date, the owner may refuse permission to view his vehicle; on several occasions we have not been allowed to inspect vehicles involved in what appear to be very low speed collisions. The police may impound a vehicle for their own investigations, and certain jurisdictions will not release their photographs and measurements. However, it often happens that the need for reconstruction does not arise until years after the accident. By this time the vehicles may long since have been repaired or scrapped, or may have been involved in another accident, and reconstruction of the accident must rely on existing photographs.

Photographing the accident scene

At the accident scene we will need to know the locations and directions of the vehicles at impact and at rest, and the location of any tire marks associated with the accident.

If the vehicles have not yet been moved they should be photographed, preferably from several directions. The photographs should include; recognisable objects such as curbs, flaws in the pavement, utility poles, trees and driveways that can later serve to re-locate the vehicles. The resting locations of pedestrians and or of occupants thrown from the vehicles are equally important.

In medium to high-speed accidents fresh gouges in the pavement typically mark the impact location, by changes in the direction of tire marks, and by debris, broken glass and fluid spills. These should be photographed from several directions, both close-up and from a sufficient distance to allow re-location in relation to surrounding objects. In low-speed accidents there may be no physical evidence of the impact area or the resting positions of the vehicles.

Tire marks leading to the impact point and from the impact to the resting position, provide important evidence of the paths of the vehicles and of their speeds. This evidence is volatile and should be photographed and measured as soon as possible after the accident. Tire marks will appear darkest when viewed along their direction of travel. The visible part of tire marks will tend to become shorter as they fade with time. Rain, or slight dampness, makes tire marks difficult to see, and if any moisture is present the scene should be re-visited and photographed again as soon as the surface is dry. Tires leave fainter marks on a wet surface, and usually none at all on a snowy, slushy or icy surface. Tire tracks in snow provide a useful record of vehicle movement, and should be photographed as soon as possible. At least one of the photographs of the tire marks should show where they lead to the impact or resting positions of the vehicles. This will help distinguish them from unrelated marks. A measurement of the "track width" (separation between left and right tires) and the tire tread width can help to identify tire marks of a particular vehicle.

A plea for good photographs

Photographs may provide the only evidence of vehicle damage, and it is important that good photographs and observations be taken as soon as possible after an accident. Photographs taken squarely from the front, back, and if possible, from above, are the best for measuring the amount of crush damage. Photographs taken facing the four corners of the vehicle can show damage that is not revealed by the lighting or angle in the "square" views. The recommended views are shown schematically in Figure 4. It is useful to record the focal length used, particularly if the camera has a "zoom" lens.

Close-up photographs of damaged areas can reveal small details such as paint transfer that are not seen in overall views. Interior views should also be photographed where there is intrusion of damage into passenger areas, or where the lack of such intrusion may be an issue. A close-up photograph of the builder’s V.I.N. plate on the doorpost confirms the vehicle identification. If safety stickers or license plates have expired they should also be photographed.

In older vehicles equipped with shock-absorbing hydraulic bumper struts, the length of scratch marks indicating the amount of compression of the struts is a valuable indicator of D V in a collision. In one low-speed accident an astute adjuster had observed, measured and photographed scratch marks on the bumper struts of a third-party vehicle. Later, we were unable to obtain permission to inspect this vehicle through the third party’s lawyer. We were able to provide a good estimate of the speeds involved in this accident on the basis of the adjuster’s measurements and photographs.

It is important that a GOOD QUALITY 35 mm CAMERA SHOULD BE USED. PolaroidÒ or "instant" cameras are popular because they provide pictures immediately. Typically they produce fuzzy pictures that cannot be enlarged to reveal details, and these pictures tend to fade and lose contrast and colour over the years. The need to finish a roll of film on one or two vehicles encourages taking an adequate number of pictures of each one. Over a year or so the initial cost of a good camera that is used daily is probably less than the cost of relatively costly "instant" film. More importantly, the cost of taking good photographs can be far less than the cost of having poor photographs or none at all.

Seat belt effectiveness

C.R. Tyner & Associates Limited personnel have investigated and reconstructed well over five hundred vehicle accidents. In almost all of these accidents we have examined the seat belts for use in the event and compared that to the outcome of the occupant. Based on that experience, we are strongly of the opinion that under almost all accident scenarios, the proper use of seat belts saves lives and reduces injury. Transport Canada has been studying the effectiveness of seat belts in reducing the probability of serious injury or death in a major Canada wide study extending over many years. Their conclusions agree with our own, that seat belts save lives and reduce the probability of serious injury. The compulsory use of seat belts has become mandatory in many countries, states and provinces. The installation of seat belts in vehicles has been mandatory in Canada since 1971.


Seat belt usage

The use of seat belts significantly reduces the probability of injury, even in low-speed accidents. In accidents where seat belt use would probably have mitigated injuries, an occupant who fails to wear a seat belt is considered to have contributed to his/her own injuries. If seat belts have been highly stressed in an accident there may be stretch marks on the webbing and/or imprints of the weave of the webbing on plastic parts of the seat belt buckle or on the D-ring that hangs the seat belt on the door pillar. In a low speed accident the forces on the seat belts are probably not sufficient to leave such marks. Even in a high-speed accident the absence of marks does not prove that seatbelts were not worn.

Signs of seat belt usage may be quite obvious, but if seat belt usage is an issue, an expert should generally do the inspection of the belts. If a vehicle must be repaired, destroyed or sold before an inspection is possible, the seat belts can be marked for identification, photographed and removed for later inspection.

Motorcycle braking and motorcycle helmets

Unlike cars, motorcycles have separate controls for front and back brakes. Generally the front brake is controlled by a lever on the right handlebar and the back brake by a pedal on the right side. If a motorcycle is not heavily damaged, the operation of the front and back brakes can be verified. The assistance of an experienced motorcyclist may be required.

In an emergency stop, because of forward weight transfer during deceleration, most of the braking power of a motorcycle is in the front brake. Motorcycle tires are made with a soft, high-traction rubber and a skillful rider can stop in a slightly shorter distance than a car by maintaining the front brake just short of skidding. In this case the front tire will leave a faint mark if any, and this mark may be hidden by a darker mark from the back tire. Locking the front wheel would quickly result in a loss of balance.

A surprising number of motorcyclists tend to use mainly or exclusively the back (foot pedal) brake in routine stopping, although in safety courses this habit is strongly discouraged. When confronted with an emergency this habit may prevail, causing a longer stopping distance than necessary. It is difficult to tell from tire marks alone whether the front brake of a motorcycle was applied sufficiently or at all, except that rear-wheel-only braking may leave a tire mark that is longer than expected in relation to the speed involved.

It is generally accepted that use of an approved helmet reduces the severity of head injuries in motorcycle accidents. Helmet use has long been required in all provinces and rarely is a motorcyclist seen without one. In a motorcycle accident there may be marks or paint transfer on both the helmet and on the object that it struck, indicating where the helmet impacted. If a helmet does not remain on the rider’s head it provides no protection and this would indicate that the chinstrap was probably not properly fastened. A motorcyclist may carry a spare helmet in case he needs to carry a passenger, in which case it is important to ascertain which helmet was actually in use.

The need to move quickly to acquire data

Accident data decays quite quickly, particularly at the scene. Figure 5 summarises the volatility of the various types of evidence discussed above. When a reconstruction is undertaken, the quality of the results may be limited by the available data. If it appears that a reconstruction might be needed we recommend that steps be taken to preserve volatile data, particularly by measuring and photographing evidence at the scene and by thoroughly inspecting the vehicles before they are repaired or scrapped. It may be worthwhile to have a reconstruction expert spend a few hours securing evidence, even if a decision to proceed with analysis and a report will not be taken until a later date.



Bailey, M., D. King and M. Gardiner, 1997: Minor Impacts. MacInnis Engineering, Vancouver, B.C.

McConnell, W.E., R.P. Howard, H.M. Guzman, J.B. Bomar, J.H. Raddin, J.V. Benedict, H.L. Smith and C.P. Hatsell, 1993: Analysis of human test subject response to low velocity rear end impacts. Biodynamic Research Corp. SAE Paper 930889.

Szabo, T.J. and J.B. Welcher, 1998: Biomechanics of low speed impacts. Paper presented at IAARS (International Association of Accident Reconstruction Specialists) Conference, Boston MA, July 1998.