We recently conducted a technical investigation into the systems modelling approach to train simulation. This blog post will cover what was created and outline the capabilities and next steps that it could be expanded upon.
Basis of the library
We used the VeSyMA library as a base, this provided the vast majority of systems that were needed. We also used the library structure and modelling approach to vehicle creation. Interfaces, templates and partial models were used to reduce duplication and improve the usability and versatility of models.
The resulting vehicle templates will look very familiar to any VeSyMA user as there is very little that is unique to trains that isn’t used in road vehicles.
Figure 1: Rolling stock base template
Above is the rolling stock base template, which, apart from some rare unusual use cases, will cover the majority of powered or unpowered rolling stock. It incorporates 2 bogies and associated rail contact models, body, coupling models, brakes and driver environment.
The couplings can be independently activated and deactivated to ensure there are no under-constrained connectors. The driver environment can also be activated / deactivated depending on whether the vehicle is directly influenced by the driver or not.
Figure 2: ICE-Electric rolling stock template
The base template was then extended to include templates of the powertrain elements, in this instance a dual motored ICE-Electric.
The only unique elements special to the railway library are the bogie and rail contact. All other elements were inherited from VeSyMA and suspensions, and used directly or slightly modified to suit application.
Additional systems can be added to these models, for example: external controllers, pneumatic systems, thermal management systems and cabin models, with the latter allowing to optimise for human comfort and energy usage. Each vehicle can be linked together with any of these systems to investigate the whole train.
Figure 3: Schematic of a multi-zone carriage HVAC system that can be fully integrated with the other vehicle systems. TiL Suite / Thermal Systems library.
Bogie, axle and rail contact
As all three elements are closely interlinked, we are addressing all at once.
The rail contact, for the initial investigations, utilised an existing joint within the Claytex library that allows 1D motion along a spline. To simplify the system this would represent an axle or bogie without the rolling rotation along the rail. It also assumes no rail deflection or loss of contact. There is a choice between a 3D rail and a prismatic for simpler longitudinal tests.
The type of rolling stock bogie suspension will influence the use of the rail contact model. If the bogie has fixed axles, where each axle is attached to the bogie rigidly, then the joint is connected to the bogie. This reduces complexity by having a single joint that controls the full motion of the bogie.
Figure 4: Bogie model with fixed axles
If there is independent suspension for each axle then they can have an individual joint for each axle and have suspension between that and the bogie.
As the rail joint does not have a 1D connector defining the length along the rail, an additional model was needed to convert the rail motion to a 1D connector to define the position along the rail. This has an optional fixed or force based friction model. The friction model acts as a simplified rail-wheel contact model, with ability for slip and loss of traction.
Figure 5: Rail contact model for fixed axle bogie
Testing
With all the components created, a couple of types of rolling stock were created, one electric locomotive and a freight style car, with no propulsion but with brakes. This was designed to test the propulsion rather than full powertrain. The electric motors were specified to appropriate levels with a single fixed ratio.
The pulled carriage brakes, that are usually pneumatic, are controlled using a real based output from the driver environment.
Figure 6: Electric locomotive
Along with all the bench tests of the sub-systems, full train experiments were created, both individual vehicles and assembled trains. The experiment will again look very similar to any VeSyMA user as it has all the same elements with the notable exception of a Rail rather than Road model.
The number of carriages is expandable, with the connect statement looped. This means back of the 1st carriage will connect to the back of the 2nd, the back of the 2nd connected to the front of the 3rd, etc.
Figure 7: Experiment layout with an expandable number of carriages
The driver model is a reduced version of the VeSyMA closed loop, as there is no lateral control and no gears, other than reverse.
The train was tested on a 3D rail with banking and turns as shown below. This particular track includes both curves and banking and is a closed circuit so can be run up to any number of cycles.
Video: Driven train around oval at 10x speed
The suspension and powertrain work as expected, with the masses of the bodies rocking on the bogies as it enters and exits the curve. In these low speed and acceleration tests there was no loss of traction but in other tests where higher acceleration is demanded the wheels start to slip.
Further Steps
While this proves the capability to create a train model within Dymola which can be used for energy analysis and systems integration modelling, using VeSyMA components, there are some areas that could be improved if the requirements drive such models to be developed. The main items that will be enhanced for future releases are the rail / wheel contact, adding lash and potentially rail deflection which will influence the torque required to move the train along the rails.
Written by: David Briant – Project Engineer
Please get in touch if you have any questions or have got a topic in mind that you would like us to write about. You can submit your questions / topics via: Tech Blog Questions / Topic Suggestion
