Simulating over 2,000 impact configurations to identify train modifications that reduce impact

Understanding injury risk

Annually, hundreds of people are hit by trains on the New York City Subway, with many of these collisions resulting in fatalities. With many large cities around the world featuring subway systems that are expanding, these incidents are gradually increasing in line with passenger numbers.

Daniel Hall conducted a study for his PhD at Trinity College Dublin to analyze these types of incidents and identify mitigation measures to protect pedestrians from serious or fatal injuries.

Hall used Simcenter™ Madymo™ software, which is part of the Siemens Xcelerator business platform of software, hardware and services, to simulate the behavior of a human body during an impact with and without countermeasures in place.

Analyzing the effect of adaptations

As little previous research has been conducted on the specific mechanisms of these injuries, they are an unknown collision type. It is unfeasible to perform physical testing, so simulation was key to the study. Hall explains that simulation enabled him to experiment with a wide variety of impact positions and velocities.

“The multibody simulation capabilities of Simcenter Madymo allowed us to create a huge test matrix of around two and a half thousand unique configurations,” says Hall.

“This meant we could look at different trends to show how adaptations would affect injury. For instance, if we were to increase or decrease the geometry of the drainage trough, how does that affect the overall head injury risk?”

These troughs are designed to allow for adequate drainage; however, empirical evidence suggests they may reduce fatalities by providing a space underneath the train for a pedestrian to fall into, so it does not run them over. However, there had not been a specific study to prove the effectiveness because there are so many factors – the type of train or track, the speed, whether it was coming around a bend, etc.

Using Simcenter Madymo, Hall created a model where he could change each variable individually to see the isolated effect.

“The best thing about Simcenter Madymo is its adaptability,” says Hall. “Being able to easily change so many things, like the position of the pedestrian, was hugely important for this study.”

The simulations showed that, although reducing speed helped with the primary impact and head injury risk, it did not result in much change in the secondary impact and amputation risk for impact velocities above 6 meters per second (m/s), which is why multibody modeling was important for the study instead of finite element (FE) modeling.

Reducing the force of impact

Hall’s final paper involved designing an energy-absorbing structure that could be mounted onto the front of the train as a countermeasure. He added this to the existing model and ran a smaller subset of the previous simulations to see how the injury risk would change with all other factors remaining the same. This analyzed primary contact with the train and secondary contact with the tracks and the risk of hitting the wheels or being run over.

“Our simulations showed that about half a meter of softer absorption material reduced the impact to a level comparable to an automotive crash,” says Hall. “When you consider the difference in weight between a train and a car this is a huge leap. A low-speed collision between a pedestrian and a car is intuitively more favorable between a pedestrian and a train due to the increased mass, stopping time and risk of being run over.”

Leveraging fast, adaptable and accurate simulation

This study proved it is possible to mitigate the impact of subway collisions. Operators now have accurate data that can influence the design of existing and future trains and better protect pedestrians against serious injury.

Ciaran Simms, professor in biomechanics at Trinity College Dublin, explains that multibody simulation was the only feasible way to carry out the study. “The alternative would be a finite element model,” says Simms. “This would have more detail, but it would also have many more unknowns, so it would produce much more uncertainty in the results. It would also take so much longer that it would be impractical to run it for so many different configurations.”

Simms highlights why a multibody simulation solution was important to the study. “Due to the complexity of the models and the volume of tests required to get a meaningful answer, using a multibody solver instead of an FE solver was critical to this study,” says Simms. “Without Simcenter Madymo, you’d need to code your own multibody modeling method, to create one such as Simcenter Madymo would take many PhDs to accomplish.”

Simms’ appreciation of the ease and speed for modeling crash mechanics with Simcenter Madymo, and how this can help students learn the complexity of crash mechanics in a visual way has resulted in him creating a course at the university.

“When I was creating a university module on collisions, I wanted to have a balance of theoretical equations and practical applications,” says Simms. “Using Simcenter Madymo was the obvious choice as it facilitates the implementation of theory and enables students to see the results. I think our minds can understand impact movements better in terms of multibody parameters than in terms of finite element models and this helps deliver more fundamental insights that can be applied to real-world scenarios.”

Although its ease of use is beneficial for learning, Hall points to the speed of the solution as an important differentiator, something that finite element models cannot match.

“When learning for the first time, you can experiment and run a simulation and know within a minute or two if it’s going to fail, which is so much better than waiting two days to find out it didn’t work,” says Hall. “I’d encourage students to just try Simcenter Madymo, and they’ll immediately see what will work and what needs to be changed. It delivers results fast, which is vital for a detailed study.”