One indicator of a motorcycle’s impact speed with a passenger vehicle is the magnitude of the translation and rotation experienced by the passenger vehicle following the impact. Simulation can be an effective means to determine the motorcycle speed necessary to cause a specific magnitude of translation and rotation of the passenger vehicle. Deyerl and Cheng [2007, 2008] illustrated this type of analysis using EDSMAC4 simulation. Another software package that can be used to carry out such simulations is PC-Crash, a vehicular crash simulation software that is widely used by the reconstruction community. Numerous studies conducted over the last two decades have demonstrated that the impact and trajectory models of PC-Crash can accurately model vehicular crashes [Rose, 2018].
In 2002, Adamson reported the analysis of seventeen crash tests that involved 1989-1993 model year Kawasaki 1000 police motorcycles. These tests were conducted at the World Reconstruction Exposition in September of 2000 (WREX2000). In the first seven tests, the motorcycles impacted a concrete block at speeds varying between 10 and 42 mph. In the remaining ten crash tests, the motorcycles impacted one of two stationary 1989 Ford Thunderbirds, both weighing just under 3,600 pounds, at various locations. Motorcycle impact speeds for these tests varied between 25 and 49 mph and were measured with a hand-held Stalker radar gun. All tests were run on an abandoned concrete runway with a reported coefficient of friction for vehicle tires of 0.72. In this and future posts, we will examine the use of PC-Crash to model and analyze these tests.
Another set of motorcycle-to-car collisions was conducted at the World Reconstruction Exposition in 2016 (WREX2016). The passenger vehicles in these tests were stationary prior to the collision. Motorcycle impact speeds varied between 30.3 and 46.3 mph. The tests were conducted on an asphalt roadway near Orlando International Airport. The coefficient of friction for this surface was measured by performing 3 skid tests with a Chevrolet Silverado and the average value was 0.7. In this and future posts, we will also examine the use of PC-Crash to model and analyze these collisions.
PRELIMINARIES RELATED TO PC-CRASH
The impact and trajectory models within PC-Crash have previously been shown to be adequate for reconstruction purposes [Rose, 2018]. Thus, the purpose of the simulations discussed here is not to validate PC-Crash. Rather, the purpose is to define valid inputs for this type of simulation and to familiarize the reader with issues that arise when simulating motorcycle-to-car collisions. For example, in simulating a motorcycle-to-vehicle collision, the reconstructionist will need to decide on a range of valid inputs for the coefficient of restitution. The simulations reported in this article will help to define such a range. In the simulations that we will discuss here, the actual impact speeds were used, and then, the simulations were optimized using the coefficient of restitution and the impact center location. This study is best characterized as a calibration study that will serve as a guide on inputs for using PC-Crash to simulate real-world motorcycle-to-vehicle collisions.
Another issue worth mentioning here is that Rose  demonstrated that some collisions induce steering at a vehicle’s front wheels during the post-impact motion. These steering angles develop, in part, because the frictional forces between the tires and ground cause the yaw rotation of the tires to lag behind the yaw rotation of the vehicle. The development of these steering angles also depends, in part, on moments generated due to the caster angle, which would cause some steering to occur even for a crash test where a significant yaw velocity did not develop. Rose demonstrated that simulations could be improved by including this steering, and so, consideration will be given to this issue in describing a simulation methodology in the sections that follow.
WREX2016, Test #3
In this first post on using PC-Crash to model motorcycle-to-vehicle collisions, we will examine one of the tests from WREX2016 (Test #3). This test involved a 2013 Harley-Davidson FXSB 103B Softail Breakout motorcycle impacting the passenger’s side front door of a stationary 2006 Nissan Maxima near its center, at an approximate 90-degree angle, and at a speed of 43.0 mph. The photograph below shows the damage to these vehicles, their rest positions, tire marks, and chalk marks identifying the passenger’s side wheel positions both pre and post-collision. As these photographs show, the front wheels of the Nissan were steered to the left at rest. Review of the video for this test confirmed that these wheels were oriented straight ahead prior to the collision and that they steered to the left during the post-impact motion of the Nissan. The motorcycle in this test weighed 667 pounds and the Nissan weighed 3,449 pounds.
In the simulation for this test, the roadway coefficient of friction was set at 0.7 and the TM-Easy tire model was used for both vehicles. Data and equations from the articles by MacInnis  and Allen  were used to estimate the moments of inertia for the Nissan. The impact height was set at 1.0 feet. A leftward steering input of 15 degrees, occurring over a time period of 250 milliseconds, was used for the post-impact motion of the Nissan. The Nissan was in park, so brake factors of 100% were used for the front wheels of the Nissan and 1% for the rear wheels. Brake factors of 100% were applied to the front wheel of the motorcycle, though the simulation was not sensitive to this input. An integration time step of 1 millisecond was used. The first image below shows the setup of the PC-Crash file with the vehicles positioned at impact. The wheel positions for the Nissan at rest are also identified in this figure. The simulation was optimized using the coefficient of restitution. A high quality match with the actual rest position of the Nissan was obtained with the actual motorcycle impact speed and a coefficient of restitution of 0.17. The video below shows the rest motion in the simulation and the match with the rest position of the Nissan.
1. Adamson, Kelley S., “Seventeen Motorcycle Crash Tests into Vehicles and a Barrier,” Society of Automotive Engineers Technical Paper Number 2002-01-0551.
2. Allen, R., Klyde, D., Rosenthal, T., and Smith, D., “Estimation of Passenger Vehicle Inertial Properties and Their Effect on Stability and Handling,” SAE Technical Paper 2003-01-0966, 2003, doi:10.4271/2003-01-0966.
3. Deyerl, Eric, Cheng, Louis, “Computer Simulation of Staged Motorcycle-Vehicle Collisions Using EDSMAC4,” white paper presented at the 2008 HVE Forum, HVE-WP2008-3.
4. Deyerl, Eric, Cheng, Louis, “Computer Simulation of Staged Motorcycle-Vehicle Collisions Using EDSMAC4,” Accident Reconstruction Journal, July/August, 2007.
5. MacInnis, Duane, Cliff, W., and Ising, K., "A Comparison of Moment of Inertia Estimation Techniques for Vehicle Dynamics Simulation," SAE Technical Paper 970951, 1997, doi:10.4271/970951.
6. Rose, Nathan A., Neal Carter, Gray Beauchamp, “Post-Impact Dynamics for Vehicles with a High Yaw Velocity,” SAE Technical Paper Number 2016-01-1470, doi:10.4271/2016-10-1470.
7. Rose, N., Carter, N., “An Analytical Review of Two Decades of Research Related to PC-Crash Simulation Software,” forthcoming through SAE, April 2018.
Nathan is an accident reconstruction expert at Kineticorp. He is dedicated to mastering his craft, and for the past 20 years, he has dedicated himself to research and writing as a means of developing authentic expertise that provides real value to juries. Nathan developed this article as part of the research that will ultimately make it into his book on motorcycle accident reconstruction.