Mind the Gap: BIM Bridges the Divide Between Old Tunnels and New Trains

Transport for London’s new Piccadilly trains are pictured here in a model. TFL determined the best design for the new trains using models of all of the line’s platforms. (Image courtesy of TFL.)

At the Bentley Year in Infrastructure 2018 Conference, one of the Reality Modelling finalist projects hit close to home, literally. Transport for London (TFL) had become a finalist for its model of the Piccadilly line.

TFL is in the process of rolling out new trains on its Piccadilly line, which will replace the 40-year-old trains currently on the line. Part of that process involved mapping out the line’s 120 platforms to ensure that the new trains would fit the old infrastructure.

According to spokespeople Pavel Srokowski and Kevin Moore, the project required an extremely precise model. The tunnels, built back in the early 20th century, were made exactly for the round tops of the old trains to save money on excavation. Any new trains would have to fit the tunnels just as perfectly, to maximize space inside the train while not scraping the tops of the tunnels. The trains would also need to fit the old platforms, and match up well with their level. There were also difficulties caused by the fact that London is an old city that has an extensive preexisting underground infrastructure.

Additionally, the Piccadilly line was initially built along the path of the above-ground roadway to avoid having to demolish buildings along the construction route. The commitment to following an above-ground path means that the tunnel has sharp turns, and that trains must compromise between building up speed on the straight sections and being able to maneuver safely around those turns.

To work with all those difficult conditions, the team generated a point cloud model of the platforms from scan data, a model the presenters said was accurate to within 20mm in all directions. They then used Dimensionally Driven Design (DDD) to analyze things like train stopping position and possible gaps between platform edges and concept train doors. Using DDD, the team was able to complete the analysis of stopping positions in a quarter of the time it would have taken if they had used more traditional methods (500 hours vs. 125 hours).

The Piccadilly line project is scheduled for completion in 2020.