EDSMAC4 Calculation Options¶

EDSMAC4 (Engineering Dynamics Company Simulation Model of Automobile Collisions, 4th Revision) is a simulation analysis of single or multiple-vehicle crashes. It is based on the program called SMAC developed at Calspan for NHTSA, and includes several major extensions and enhancements developed by Engineering Dynamics Company, LLC. EDSMAC4 uses a set of assumed or estimated initial conditions, including positions, velocities and driver controls, provided by the user for each vehicle. EDSMAC4 then predicts the outcome of a crash by simulating the vehicle trajectories and damage profiles.

The output generated by EDSMAC4 is a time history of the vehicle kinematics (position, velocity and acceleration), kinetics (summation of tire and collision forces and moments acting at the CG), and tire skidmarks for the duration of the crash sequence. The program also computes the resulting damage profiles, the Collision Deformation Classifications (CDC), Principal Direction of Force (PDOF), delta-V (a measure of crash severity) and peak acceleration for each damage range. EDSMAC4 also produces a collision pulse (acceleration vs time history) useful for occupant simulations, such as EDHIS.

Accident investigators can use EDSMAC4 to determine how the crash may have occurred. By repeated adjustments of the initial conditions and driver braking (or acceleration) and steering inputs, the user will converge on the data that best match the known accident site evidence (usually rest positions, skidmarks and vehicle damage).

In addition to vehicle-to-vehicle collisions, EDSMAC4 may also be used for many types of barrier collisions, including pole impacts.

Users of EDCRASH will find it convenient to first analyze the accident using EDCRASH to estimate the impact speeds, then use these speeds as the initial conditions for the EDSMAC4 program. The resulting confirming analysis is very powerful.

Accident researchers can use EDSMAC4 to simulate an actual staged collision test. This can be done before the test in order to predict or anticipate the result and help during test preparations. Test parameters may be adjusted based on EDSMAC4 results to help ensure the test is successful. After the test, the researcher can note the differences between the actual and simulated collision and modify the EDSMAC4 input in order to obtain a better match.

An extremely useful feature of EDSMAC4 is the ability to quickly and easily review the results associated with different input scenarios. Termed what if analysis, changes can be made to an isolated variable or set of variables in order to identify their effect on the outcome. Only the inputs that change need to be modified; the program saves all the previously entered information. For example, the sensitivity of the result to initial velocity, or post-impact driver control, can be studied by merely changing those parameters.

Another useful feature of the program is the ability to display the results in 3-D viewers. These trajectory simulators display the vehicles at user-specified time intervals during the sequence. Using HVE's video interface, it is simple and easy to route the trajectory simulation to video tape.

EDSMAC4 has the following parameter options, listed in dialog order. Each is stored with the event and read by the physics engine when the event is executed.

Vector Spacing (deg)¶

Default: 2.0 degrees.

This angle determines the angular interval between the RHO vectors that define each vehicle's crush perimeter. (The old help recommended a value of 1 degree; the current program default is 2 degrees.)

Vector Adjustment Increment (in)¶

Default: 0.20 in.

This increment determines the incremental adjustment for each RHO vector as the collision routine seeks to establish force equilibrium between vehicles.

Vector Force Tolerance (lb/in)¶

Default: 15.0 lb/in.

This value is the allowable difference in the force computed along corresponding RHO vectors for each vehicle. A small value is obviously preferable. In general, it should not exceed 50 lb/in (about 100 N/cm).

Min Velocity for Friction (in/sec)¶

Default: 5.0 in/sec.

The minimum relative tangential surface velocity at which the full inter-vehicle friction is achieved. Below this velocity the friction force scales linearly with sliding velocity: the effective friction coefficient becomes \(\mu\,(v_t / v_{min})\), where \(\mu\) is the Inter-Vehicle Friction value, \(v_t\) the relative tangential velocity and \(v_{min}\) this minimum velocity.

Note: the old on-line help interchanged the descriptions of this option and inter-vehicle friction. The description given here is correct: this field is the minimum velocity, not the friction coefficient.

Restitution Constant¶

Default: 0.04606.

This value is the constant term used for the parametric restitution model used in EDSMAC4.

Restitution Linear Coef (1/in)¶

Default: 0.0017547 1/in.

This value is the linear term used for the parametric restitution model used in EDSMAC4.

Restitution Quadratic Coef (1/in^2)¶

Default: 0.000016711 1/in².

This value is the quadratic term used for the parametric restitution model used in EDSMAC4.

Inter-Vehicle Friction (removed from this dialog)¶

The Inter-Vehicle Friction coefficient (default 0.55) is no longer set in the EDSMAC4 Calculation Options dialog. It is now assigned per vehicle pair in the Vehicle Mesh dialog (Inter-Vehicle Options group), and the physics engine takes the collision friction coefficient from that setting.

Use Vehicle Geometry for Rho Vector Init¶

Default: unchecked (off).

When checked, the initial lengths of the RHO (crush perimeter) vectors are computed from the vehicle's actual body mesh geometry rather than from the idealized rectangular vehicle outline. This improves collision detection and damage profiles for vehicles whose plan-view shape departs significantly from a rectangle.

Steer DOF¶

Default: Off. Choices:

Off (default) — The steer degree of freedom is not simulated; wheel steer angles follow the driver steer tables directly.
Normal — The dynamic steer degree of freedom is active for the entire event; steer angles are computed from the torques acting on the steering system (wheel friction, steering stops, gyroscopic precession, column friction) and the driver steer tables are ignored.
Append — The steer tables are followed to their end; the final table entry then becomes the initial condition for the steer degree of freedom, which controls the steering for the remainder of the run.
AutoStart — The steer tables are followed until the first collision contact; at impact the steer degree of freedom starts automatically with the current steer angle and zero steer velocity. Useful for modeling a driver's loss of steering control at impact.

Inter-Vehicle Connection¶

Default: Adjust Both Equally. When a tow vehicle and trailer are connected, the heights of their user-entered hitch points may not match. If the height mismatch exceeds 0.1 in, this option selects how the gap is closed at event initialization — by adjusting the connection points' z-coordinates on the vehicles, not by moving the vehicles themselves:

Manual (User-Adjusted) — No automatic adjustment; the user must enter connection heights that match.
Adjust Tow Vehicle — The tow vehicle's rear connection height is adjusted to meet the trailer's.
Adjust Trailer — The trailer's front connection height is adjusted to meet the tow vehicle's.
Adjust Both Equally (default) — Each connection height is adjusted by half of the mismatch.

Connection Model¶

Default: Use Heavier Vehicle. Selects how the connection (hitch) constraint stiffness and damping are derived when computing the forces between connected vehicles:

Use Both Vehicles — Spring rate and damping are derived from the lighter of the two vehicles' weights.
Use Heavier Vehicle (default) — Spring rate and damping are derived from the heavier vehicle's weight.
Use Tow Veh Properties — The tow vehicle's user-entered hitch stiffness and damping values are used directly.

Hydroplane Model¶

Default: Off. Choices:

Off (default) — No hydroplaning effects.
NASA — The NASA (Horne/Borne) hydroplaning model is applied at each tire travelling over a water polygon. The model predicts the hydroplaning (spin-down) speed from the tire's inflation pressure alone: a tire hydroplanes when its forward ground-plane speed exceeds \(182.16\sqrt{P}\) in/sec, with \(P\) in lb/in² (equivalently \(10.35\sqrt{P}\) mph). Above that speed the tire's friction multiplier switches from the surface friction to the friction assigned to the water polygon, sharply reducing the available longitudinal and lateral tire forces.

The friction multiplier is applied to both the peak and sliding friction coefficients taken from the tire's friction tables, so hydroplaning reduces the tire force available longitudinally and laterally — braking, drive traction and cornering capability are all lost together.

NOTE: The friction reduction described above is the whole of the model's effect. The extra drag of displacing water is not modeled.

(HVE also provides NASA-TTI and Gallaway hydroplaning models — both offered by SIMON, see SIMON — Hydroplaning Model — and a Blythe-Day model that is not implemented in any program. Only Off and NASA appear in the EDSMAC4 dialog; any other selection is rejected with a fatal error and the event will not run.)

Accident History Basis¶

Default: Impact Force. Selects the criterion used to detect the beginning and end of each collision phase reported in the Accident History (damage ranges, delta-V, peak acceleration):

Impact Force (default) — A collision phase is bounded by the presence of inter-vehicle collision force. Impact is declared on the first timestep on which collision force acts; separation is declared one timestep after it ceases.
Acceleration — A collision phase is bounded by the vehicle's total acceleration exceeding the Threshold (g) value. The threshold edit field (default 1 g, initialized from the environment's gravity) is enabled only when this basis is selected. Impact is declared when either vehicle's total acceleration exceeds the threshold; separation is declared after both vehicles have remained below it for six consecutive timesteps.

In both cases the vehicles' perimeters must also overlap before impact can be declared — the acceleration test is an additional condition, not a substitute for contact.

NOTE: The acceleration used by the Acceleration basis is the vehicle's total acceleration, which includes tire forces. A threshold low enough that hard braking or cornering exceeds it on its own can prevent separation from ever being declared, leaving the event in the collision phase — and at the collision timestep — for the remainder of the run.

NOTE: The Threshold value is used by the Traditional Damage Data Format regardless of which basis is selected. Although the field is disabled when the basis is Impact Force, the stored value (1 g by default) still sets the minimum acceleration at which an acceleration peak is recognized and delta-V accumulated in that report.

See EDSMAC4 — Impact and Separation Times for the full description.

Damage Data Format¶

Default: Collision Data. Choices:

Traditional — Damage is reported in the traditional EDSMAC-style damage-profile format: damage ranges along the crush perimeter, each with a CDC, PDOF, total delta-V, time of peak acceleration and peak acceleration. The severity figures are derived from peaks in the vehicle's acceleration history.
Collision Data (default) — Damage is reported using HVE's extended Collision Data format: one row per collision pulse, with the vehicle struck, the pulse start and end times and duration, peak acceleration, peak collision force, delta-V and PDOF, followed by separate damage-profile and crush tables. The severity figures are derived from the inter-vehicle collision force and the accumulated velocity change.

This option selects the calculation as well as the layout. The two methods are independent and will not give identical PDOF, delta-V or peak acceleration values; see EDSMAC4 — Collision Severity Results. The Traditional format is retained for comparison with results from earlier releases, and is only available for events with more than one vehicle.

Include Free Space — When checked, the reported crush measurements include free-space (induced) deformation in addition to direct-contact crush. In the current release this checkbox is permanently checked and disabled; free space is always included.

Terrain Search Options (Get Surface Information)¶

The old help ended with three undocumented entries — From First Polygon, From Previous Polygon and By Elevation. These options select how HVE searches the terrain polygon (triangle) database beneath each tire at every timestep. They are no longer part of the EDSMAC4 Calculation Options dialog; they are set in the separate Get Surface Information Options dialog (Options menu), and the Search Method, Search Direction and z-component settings made there are stored with each event. The behavior is:

Search Method (default: From Previous Polygon, Sorted):

From First Polygon — The entire terrain database is searched in order starting from the first polygon; the first polygon containing the search point is used. Simple but slowest for large terrains.
From Previous Polygon — The search starts at the polygon found on the previous timestep and spreads outward (forward and backward) through the database, returning on the first hit. Fast, but friction-zone precedence is not enforced (water polygons are still checked first).
From Previous Polygon, Sorted (default) — Water, curb and friction-zone polygons are checked first (so friction zones take precedence over the underlying road surface), then the search proceeds from the previous polygon, spreading outward. This is the recommended method.
By Elevation — Intended to search the entire database and return the polygon with the highest elevation beneath the wheel center (for overlapping surfaces such as bridges). Not currently implemented: the radio button is disabled in the dialog, and if the option is ever selected the physics engines report a fatal error and the event will not run.

Search Direction (default: Upward Facing Only) — Controls which terrain triangles are considered usable when the terrain database is built:

All Directions — All triangles are used regardless of the direction of their surface normal.
Upward Facing Only (default) — Only triangles whose surface normal has an upward component are used (downward-facing and vertical triangles are discarded). Curb polygons are exempt from the normal test.
Z Component, Greater Than — Only triangles whose unit surface normal has a z-component greater than the user-entered value (default 0.0) are used. The edit field is enabled only for this choice.

DyMESH¶

EDSMAC4 does not use the DyMESH 3-D collision algorithm; the DyMESH Options dialog applies to SIMON events only.

Updated to match the current version of HVE.

← Previous: EDSMAC Calculation Options | Index | Next: EDSVS Calculation Options →