Tools › Mechanical Design › Vehicle Dynamics › Accident Reconstruction


Accident Reconstruction

The standard speed, energy and time-distance calculations of vehicle collision reconstruction, each reported as a low, nominal and high value from the entered ranges of friction and measurement: speed from skid marks over several surfaces with grade and braking efficiency, critical speed from a yaw mark by chord and middle ordinate with superelevation, launch speed for a fall, flip or vault, impact speeds and speed changes of two vehicles by conservation of momentum in the plane, crush energy and equivalent barrier speed by the NHTSA CRASH3 damage method with its stiffness categories and force-direction correction, and time and distance before impact with perception-reaction time and an avoidance check. Each result is shown as formula, numbers and answer. The tool reports quantities only and draws no conclusion about fault.

  • Skid to stop or to impact, critical speed yaw, and fall, flip or vault speeds, each with a low, nominal and high value

  • Planar momentum for two vehicles: impact speeds, speed changes, energy lost, with the in-line case treated separately

  • CRASH3 crush energy from two, four or six crush depths, equivalent barrier speed and two-vehicle speed change

  • Time and distance before impact: braking, perception-reaction, where the vehicle was, whether it could have stopped

Accident Reconstruction


Unit System: /

/

Analysis


Skid Marks, in the Order Traveled (friction as a low and a high value)

1
2
3
4

Skid Conditions




Yaw Mark





Fall, Flip or Vault





Momentum: Vehicles, Headings and Post-Impact Speeds











Crush: This Vehicle (depths left to right across an end, rear to front along a side)
























Crush: Other Vehicle (entered energy in ft-lb in USCS, J in SI)























Time and Distance: Vehicle, Hazard and Other Party












Friction Range



Measurement Ranges (plus and minus)





×

Accident Reconstruction

Accident Reconstruction: Methods, Ranges, Assumptions and Limits

What the tool does

The tool carries out the standard hand calculations of vehicle collision reconstruction: speed from skid marks, critical speed from a yaw mark, launch speed for a fall, flip or vault, impact speeds from conservation of momentum, crush energy by the NHTSA CRASH3 damage method, and time and distance before impact. Each result is shown as formula, numbers and answer, and then as a low, nominal and high value that follows from the ranges entered for friction and for the measurements. The tool reports quantities. It does not say who or what caused a collision.

How the ranges are evaluated

Friction and reaction time are entered as a low and a high value, and their nominal value is the middle of the range. Lengths, angles, entered speeds and crush stiffness take a plus-and-minus tolerance about the entered value. The tool evaluates the result at the nominal entries and at every combination of the low and high ends of the ranged entries (2k combinations for k ranged entries), and reports the least and the greatest result. For the formulas here, which rise or fall steadily with each entry over the usual ranges, these are the true bounds for entries inside the ranges. They are bounds, not a probability statement: the chance that every entry sits at its unfavorable end together is small, and a statistical treatment would give a narrower interval. Where a result does not rise or fall steadily with an entry (the momentum solution near in-line headings, the force angle of the crush analysis near zero) the corner values can miss an extreme that lies inside the range; the nominal point is always included.

Skid to stop or to impact

A vehicle sliding with a drag factor f loses speed at the rate f g. Over a distance d the work done equals the kinetic energy lost, so v0 = √(ve² + 2 g Σ fi di), where ve is the speed at the end of the marks (zero for a stop) and the sum runs over the surfaces crossed. In customary units with one surface this is the familiar S = √(30 d f) in miles per hour and feet. The drag factor of a segment is f = μ η + G: tire-road friction coefficient μ times braking efficiency η, plus the grade G as a decimal (positive uphill). The equivalent single drag factor Σ f d / Σ d is the form given in the CRASH3 Technical Manual (page 1.92).

What it assumes. Constant friction on each surface, a point mass, all of the listed distance traveled at the stated efficiency, and the small-slope form of the grade term (the exact form μ cos θ + sin θ differs by less than one percent for grades to 10 percent). When it misleads. The result is the speed where the visible marks begin, a lower bound on the speed before braking: speed is lost before tires mark the road, and anti-lock brakes may leave faint marks or none. Friction taken from a table can differ from the surface on the day; test skids at the scene are better. A vehicle that was struck, steered or rotated during the slide does not fit the straight-skid model.

Critical speed yaw

The radius of a circular arc follows from a chord c and its middle ordinate m: r = c²/(8 m) + m/2. A vehicle rounding a curve of radius r on a road with superelevation e (as a decimal) reaches the friction limit at v = √(g r (μ + e)/(1 − μ e)), from the balance of forces on a banked surface. With e = 0 this is v = √(μ g r). When a track width is entered, half of it is taken from the radius of the tire mark to give the path of the center of mass.

What it assumes. A true yaw mark: tires rolling and slipping sideways at the limit of friction, no braking or acceleration, a constant radius over the measured chord, a point mass. When it misleads. The radius is very sensitive to the middle ordinate: on a 50 ft chord, an error of one inch in a 1.5 ft ordinate changes the radius by more than 5 percent. Marks from a braking or spinning vehicle, or a chord taken late in the mark where the vehicle has slowed and the radius has tightened, give a speed that is not the speed at the start of the yaw.

Fall, flip and vault

The center of mass of a vehicle that leaves the ground follows a parabola. Leaving at speed v and angle θ above the horizontal, traveling a horizontal distance d and landing h lower, v = (d / cos θ) √(g / (2 (d tan θ + h))). For a level take-off (θ = 0) this is the fall formula v = d √(g / (2 h)). When the angle is unknown, as in a flip or vault, the least speed that reaches the landing point is vmin = √(g (√(d² + h²) − h)), at the angle θ = ½ arctan(d/h); on level ground this is 45° and v = √(g d). The customary flip and vault formula assumes 45° even when the landing is lower or higher, v = d √(g / (d + h)); the tool prints it beside the true minimum.

What it assumes. No air resistance, no loss of speed at take-off, distances measured for the center of mass between the take-off point and the first contact. When it misleads. The speed with an unknown angle is a minimum only. A vehicle that tripped or struck a curb lost speed at take-off that the formula does not see. Distances taken to the final rest position rather than to first contact overstate the speed.

Planar momentum

For two vehicles of mass m1 and m2, the momentum just before impact equals the momentum just after it. With approach headings α1, α2, departure headings β1, β2 and post-impact speeds v3, v4, the two component equations give

v1 = (Px sin α2 − Py cos α2) / (m1 sin(α2 − α1)), v2 = (Py cos α1 − Px sin α1) / (m2 sin(α2 − α1)),

where Px and Py are the components of the total momentum after impact. The speed change of each vehicle is the length of the vector difference between its velocity after and before impact. The weight entered is the mass of the vehicle as loaded: in customary units the scale weight in pounds is used directly, since only ratios of mass enter the speeds.

The in-line case. Both speeds are divided by sin(α2 − α1). As the approach headings near a common line (head-on or rear-end) this divisor goes to zero and small errors in the angles are magnified without limit: the two equations become one. The tool prints the magnification, warns within about 15° of a common line, and offers the in-line analysis, in which one impact speed is entered from other evidence (zero for a stopped vehicle) and the other follows from momentum along the approach heading. The momentum left unbalanced across that heading is printed as a check on the entries.

What it assumes. Forces from the tires and the road during the short impact are negligible beside the collision force; the masses are known; the post-impact speeds and directions are those of the centers of mass at separation. When it misleads. Post-impact speeds usually come from slides to rest whose friction is itself uncertain, so their error carries straight into the result. A heavy vehicle against a light one makes the light vehicle's impact speed poorly determined. Results that show a negative impact speed, or more kinetic energy after the impact than before, mean the entries do not fit together.

Crush energy (CRASH3)

The damage method of the NHTSA CRASH3 program, after Campbell, takes the force per unit width needed to crush a vehicle as A + B C, a straight line in the residual crush depth C. The energy absorbed over a damage width L is the integral of A C + B C²/2 + G, with G = A²/(2B) the energy stored before permanent crush begins (CRASH3 Technical Manual, equations 2.46 and 2.47). For two, four or six equally spaced depths joined by straight lines the integral is the manual's equations 2.50, 2.49 and 2.48, which the tool evaluates in the single form E0 = (L/(n − 1)) Σ [A (Ci + Ci+1)/2 + B (Ci² + Ci Ci+1 + Ci+1²)/6 + G].

The coefficients describe crush perpendicular to the damaged side or end. The force direction is entered as CRASH3 defines it, a clock direction on the vehicle (0° from straight ahead, positive clockwise), together with the damaged surface; the angle α between the force and the perpendicular to that surface follows. The energy is multiplied by 1 + tan² α, with α limited to 45° (equation 2.58 and the discussion that follows it in the 1986 manual, page 2.38). The earlier User's Guide and Technical Manual (page 9-18) limits the correction to 75°; in Advanced mode the limit is an input that accepts up to 75°. The equivalent barrier speed is EBS = √(2 E / m). With the crush energy of both vehicles, ET = E1 + E2, the speed changes are (equations 2.44 and 2.45)

ΔV1 = √(2 γ1 ET / (m1 (1 + γ1 m1/(γ2 m2)))), ΔV2 = √(2 γ2 ET / (m2 (1 + γ2 m2/(γ1 m1)))),

with γ = k²/(k² + h²), where k is the yaw radius of gyration and h the moment arm of the collision force about the center of mass; γ = 1 for a central impact.

Moment arm. By default the arm is computed as the CRASH3 program forms it (User's Guide and Technical Manual, section 11.3, pages 11-6 to 11-9): the force passes through the centroid of the damaged area along the entered force direction, and the arm is the perpendicular distance from the center of mass to that line. The centroid lies xc along the surface from the middle of the damage width and dc in from the original surface, both found from the crush profile (dc = ∫ (C/2) C dl / ∫ C dl). With the distance from the center of mass to the damaged surface (XF, XR or YS) and the offset D of the damage midpoint, the centroid is at (XF − dc, D + xc) for front damage, (−(XR − dc), D + xc) for rear damage and (D + xc, ±(YS − dc)) for the right and left sides, in axes x forward and y to the right; for a force from the clock direction p the arm is h = |y cos p − x sin p|. Crush depths are entered from left to right across an end and from rear to front along a side. The computed arm is the default because it is formed again at every corner of the entered ranges: a tolerance on the force direction then moves the arm, and through it the share of the mass that acts at the contact, as well as the energy correction. An entered arm stays fixed over the range. The other vehicle's energy and arm are computed in the same way when its damage is entered; the angle tolerance turns the one line of force on both vehicles together.

Vehicle dimensions. The size-category select fills the three distances and the radius of gyration from Table 8-1 of the User's Guide and Technical Manual (categories 1 to 6 by wheelbase). The table gives default dimensions for passenger cars of the 1970s and early 1980s; the fields stay editable, and the dimensions of the vehicle concerned should be entered where they are known. The van column of the table is not offered because one of its cells is illegible in the available copy.

Stiffness categories. The select offers the A, B and G values of Table 1.2 of the manual by vehicle size and damaged end. The manual states that they rest on vehicles of the 1970s. They are offered so that the manual's method and sample run can be followed; coefficients fitted to crash tests of the vehicle concerned should be entered where they exist. The manual prints G to the nearest pound and its values differ slightly from A²/(2B); the tool uses the printed G for a category and A²/(2B) for entered coefficients.

What it assumes. A linear force-crush law, crush uniform over the height of the vehicle, a common velocity reached at the contact area, no sliding between the vehicles, and no restitution beyond what the fitted coefficients carry. When it misleads. The manual itself notes reduced accuracy in low-speed collisions (page 2.22), where stiff parts such as wheels meet, and in side impacts where the sill is overridden (page 2.25). The 1 + tan² α factor doubles the energy at 45°, and the manual records the criticism that it is too large (page 2.37). The equivalent barrier speed is neither the impact speed nor the speed change. The tool follows the printed equations and the printed coefficient table; the energies printed by the program in the two documents' sample runs differ from it by up to about 0.4 percent, well inside the uncertainty of the stiffness coefficients. In the sample run of the User's Guide (pages 5-8 and 5-9) the program's printed energy for the first vehicle, struck at 45°, omits the direction correction that both documents describe; the tool applies it.

Time and distance

Working back from the impact: a vehicle that braked over a distance db at drag factor f and struck at vi was traveling at v0 = √(vi² + 2 f g db) when braking began, tb = (v0 − vi)/(f g) before impact. Its driver perceived the hazard one perception-reaction time tr earlier, at a distance db + v0 tr from the impact point. The position at any earlier time follows from the same two phases. The total stopping distance from v0 is v0 tr + v0²/(2 f g). The greatest speed from which the vehicle stops within a distance D is the root of v tr + v²/(2 f g) = D, v = f g (−tr + √(tr² + 2 D/(f g))). For another party the time to cover a distance d is d/v at constant speed or √(2 d/a) from rest at constant acceleration.

What it assumes. Constant speed before braking, a constant drag factor from the start of braking, a single perception-reaction time, straight travel. The opening reaction-time range of 0.75 to 1.5 s is the normal range stated in the FHWA guide Speed Concepts (page 10), which also gives the 2.5 s used for highway design. When it misleads. Reaction time depends on expectancy, lighting and the conspicuity of the hazard and can lie outside any standard range. The point at which a hazard could first be perceived is a judgment, not a measurement. The avoidance comparison holds everything else fixed; it does not consider steering, or that at a lower speed the other party would have moved on by the time the vehicle arrived.

Friction values

Friction is an entered range. The opening values, 0.55 to 0.70, are those the CRASH3 Technical Manual (Table 1.3, reproduced there from Baker's Traffic Accident Investigation Manual, 1975) lists for traveled dry asphalt at more than 30 mph. The table itself is not reproduced here. Wet, icy, gravel and polished surfaces are much lower, and modern tires and anti-lock braking on dry pavement can be higher.

Not included

Pedestrian throw-distance formulas are not included: no formula with a stated range of validity was found in a public document. Also not included: derivation of stiffness coefficients from crash tests, a range on vehicle dimensions and on the damage offset, restitution, vehicle rotation after impact, rollover, motorcycle and articulated-vehicle methods, and statistical (Monte Carlo) treatment of the ranges.




References

National Highway Traffic Safety Administration (1986). CRASH3 Technical Manual. Accident Investigation Division, National Center for Statistics and Analysis, U.S. Department of Transportation, Washington, DC.

National Highway Traffic Safety Administration (1982). CRASH3 User's Guide and Technical Manual. Publication DOT-HS-805732, U.S. Department of Transportation, Washington, DC.

McHenry, R. R. (1976). Extensions and Refinements of the CRASH Computer Program, Part I: Analytical Reconstruction of Highway Accidents. NTIS Report PB-252114, Calspan Corporation, Buffalo.

Campbell, K. L. (1974). "Energy Basis for Collision Severity." SAE Technical Paper 740565, Society of Automotive Engineers, Warrendale.

Donnell, E. T., Hines, S. C., Mahoney, K. M., Porter, R. J., and McGee, H. (2009). Speed Concepts: Informational Guide. Report FHWA-SA-10-001, Federal Highway Administration, Washington, DC.

Baker, J. S. (1975). Traffic Accident Investigation Manual. The Traffic Institute, Northwestern University, Evanston.

Fricke, L. B. (1990). Traffic Accident Reconstruction. Volume 2 of the Traffic Accident Investigation Manual, Northwestern University Traffic Institute, Evanston.

American Association of State Highway and Transportation Officials (2018). A Policy on Geometric Design of Highways and Streets, 7th ed. AASHTO, Washington, DC.

Worked Examples

Published examples this tool reproduces; each page lists the inputs to enter and the printed answers.

All worked examples


DISCLAIMER: The methods and calculation on aNOWysis.com are provided as means to educate the reader. There is no warranty, expessed or implied, as to the relevance or accuracy of the methods or calculations. All calculations should be checked by a qualified analyst for appropriateness and accuracy. We will not be held liable for any damages, injuries or deaths caused by using methods or calculations from this site.