Simulating bicycle crashes to determine the body parts that would benefit most from protective equipment

Seeing that cycling injuries are increasing

Ireland experienced a significant increase in cycling in the 2010s, driven by both the government promoting more sustainable travel and individuals’ desire to get healthier. This increased the number of cyclists experiencing impacts, which resulted in injuries. However, not all these impacts involved motor vehicles. A significant proportion of incidents were people falling off bikes without the involvement of another road user.

Having found similar data in international research, Kevin Gildea decided to carry out his PhD study at Trinity College Dublin by investigating these types of collisions and the injuries cyclists incur when impacting the ground.

Gildea used Simcenter™ Madymo™ software, which is part of the Siemens Xcelerator business platform of software, hardware and services, to simulate the behavior of the human body during these incidents and identify the greatest injury risks.

Simulating a fall accurately

When someone falls off a bicycle, they instinctively put out their arms to try to protect themselves from the impact of hitting the ground. Therefore, it was important to accurately replicate this in the simulation study. As there had not been much previous investigation into this type of fall, existing methods could not easily incorporate this type of motion. Thus, Gildea captured video footage of people falling off bicycles and extracted the motion to map to a computational model.

Gildea explains why he chose to use multibody simulation instead of a finite element (FE) model. “It’s extremely difficult and time-consuming to reposition a finite element model,” says Gildea. “There are entire PhD projects dedicated to having a model adopt a specific pose, but we wanted something more easily configurable that would give us time to answer the problem we were trying to address.”

Ciaran Simms, professor in biomechanics at Trinity College Dublin, says the solution they chose gave the ideal starting point to work from.

“The great thing about using Simcenter Madymo is not only the rapid multibody solver but that it also has a library of human models that have been comprehensively validated and used by researchers and the industry for decades,” explains Simms. “Rather than starting from scratch, we had an excellent baseline to work from. Instead of spending time creating and validating models, we could use Simcenter Madymo to get straight to the problem we want to address.”

“Leveraging the efficiency and adaptability of Simcenter Madymo allowed us to easily set conditions using the joint orientations and initial velocities and accelerations,” says Gildea. “This meant that once we extracted the motion data from the video footage, configuring the model was straightforward.”

Identifying injury risk

The most common form of protective gear for cyclists is a helmet. However, the data showed that less than 20 percent of these types of falls resulted in any impact to the head. The ability to map joint positions and accelerations in Simcenter Madymo models and then run the simulation to obtain the injury results accurately replicated the incident, including the instinctive human nature to protect oneself. The results showed injuries were most likely to be distributed across various limbs, as the individual used them to catch themself or protect vital areas.

“To accurately model the injury, we had to stiffen the joints of the limbs just like how a person would resist their motion with their arms to prevent an impact to their head,” explains Gildea. “Using Simcenter Madymo XML files were ideal for this, since we could open them with a simple script, adjust parameters and initiate the solver.”

To find the correct stiffness of the joints, Gildea used a generic algorithm with a population of 200 that ran for eight generations.

“Being able to use scripts to easily edit the parameters and the speed of the Simcenter Madymo solver were essential to this project being possible within a reasonable timeframe,” says Gildea.

With a model that more accurately replicates human motion, Gildea could determine that injuries to the head were rare, whereas injuries to the extremities were far more common. This is a novel method of taking widely available video footage to better understand the mechanisms and likely rates of injury.

Influencing the design of protective gear

This study highlights the areas where cyclists are most likely to suffer injury and therefore the type of protective equipment they need. Head injuries do still occur, but it shows exactly which part of the head is most at risk; thus, helmets can be designed with this in mind. It also feeds into the research and design of thorax airbags – wearable vests that inflate during an incident to protect the body.

The same methods are now being used for related studies where it is crucial to understand human body behavior during a fall. These include competitive cycling, where there are often falls involving multiple people, and equestrian sports, where injuries can be particularly severe due to the long fall from a horse to the ground.

This Trinity College Dublin study was funded by the Road Safety Authority (Ireland).