CNR Qiqihar Railway Rolling Stock
China's railways are among the busiest and largest in the world. More than 110,000 kilometers of track connect all major cities and industrial areas for transporting passengers and goods.
Designing robust railway vehicles
Transporting freight by train can be more energy efficient than by trucks or ship, especially over long distances. China annually transports over 4 billion tons of bulk cargo by rail, primarily coal, ores, minerals and grain. Driven by the enormous growth of its heavy industries, China plans to expand the network to 270,000 kilometers of track by 2050 to accommodate the expected increased demand for freight traffic capacity.
These rolling stock body structures undergo severe dynamic loading while in operation. The roughness of the rails, braking, curve passing and other cyclic or transient events subject them to enormous repeated stress over their lifetime. This can cause cracks and fractures in components such as the body, bogies, axles and bolsters. Therefore, avoiding fatigue damage is of primary importance when designing freight cars. On top of that, the market continuously demands faster transport in lighter vehicles to reduce time and cost. This could represent contradictory requirements, as higher speeds increase the dynamic loads, while reducing the amount of material could weaken the structure.
Railway engineers spend a substantial amount of development time trying to meet their customers’ financial considerations without compromising safety and durability. As so-called “design by experience” leads to recurrent prototype testing and long and expensive development cycles, QRRS experts are constantly looking for new durability engineering technologies, including simulation, that frontloads design decisions and delivers better performance on the first attempt, drastically shortening development time.
Efficient and reliable rail-load predictions
Crucial to successful fatigue life predictions is defining correct rail loads and understanding how they affect individual wagon components. Calculating them is challenging, especially because of the complex friction relation between the wheel and rails. Measurement data is usually either only partially available or not available at all during the concept phase.
Over the years, multibody simulation has been promoted as a reliable technology for accurately calculating rail loads. Several methodologies exist to handle different levels of model complexity and available measured data. Product lifecycle management (PLM) specialist Siemens PLM Software has developed a full computer-aided engineering (CAE) approach called “the digital test track,” as well as a combined test and CAE approach. The latter, included in LMS Virtual.Lab™ Motion software, is based on time waveform replication (TWR), originally a test technology for exact recreation of measured field data on a laboratory shaker system.
This process provides QRRS engineers with an efficient and reliable durability engineering approach. It gives them full insight into the dynamic behavior of all wagon components and helps them make the right decisions in the early stages of design.
“Noise, vibration and harshness (NVH), and especially its effect on durability, are critical design criteria for us,” says Wen Wang, team leader of simulation and analysis at QRRS. “So we want to consider them as early as possible in the vehicle development process. But at the concept stage of our design, there is obviously no prototype available and we simply cannot just get the loads from a rail test or a test rig.
“Simulation is the only option. But in the software packages we used, we had to create a digitized rail and needed to model the complex interaction between the rails and the wheel. This is very time consuming and difficult to do.
“The experts from Siemens PLM Software came up with a very pragmatic approach. The solution in LMS Virtual.Lab Motion allows us to start from measured displacements on an existing vehicle and calculate equivalent input signals for a new model. They can be applied to a fully detailed multibody model that simulates the dynamic response. These results include the loads on the individual components for NVH analysis or fatigue-life prediction, just like we wanted.”
Adopting the TWR approach
The TWR-based approach can start from any type and number of tested data. The process replicates a laboratory vehicle rail load test using an unconstrained multibody model and experimental data measured on a similar existing structure as boundary condition. Through an iterative control technique, back-calculated equivalent drive signals can be computed. By computing these drive signals for the wheel centers, LMS Virtual.Lab Motion doesn’t need to be used to model complex elements, e.g., friction models between the wheels and rails, and digitized rails.
Moreover, it eliminates the risk that the body model includes redundant constraints. The simulation corresponds to a durability test rig. LMS Virtual.Lab Motion features a powerful and accurate solver and dedicated modeling functionalities such as flexible bodies, as well as specific elements for fatigue analysis on spot and seam welds. The software then enables the user to calculate the individual component loads for use in a simulation with LMS Virtual.Lab Noise and Vibration software or LMS Virtual.Lab™ Durability software.
“Together with the experts from Siemens PLM Software, we validated this technology on one of our existing vehicles,” says Wen Wang. “Starting from a detailed multibody model and measured accelerations on wheel axle, body and certain points on the bolster, we calculated an equivalent displacement excitation that simulates a 12-channel virtual bench. That includes eight vertical and four horizontal channels for the front and rear bogie. We used this excitation for a forward dynamic response calculation.”
The results from LMS Virtual.Lab Motion corresponded well to the measurements in some observation points, both in time and frequency domains.
“We could see that the vertical and lateral response of the body and bolster were accurately simulated,” confirms Wen Wang. “The correlation coefficient between measured and calculated data approached one. And even though the model includes some complex elements like nonlinear bushes, springs and friction factors, it only took a limited number of iterations until we had convergence when calculating the loads. This gave us full confidence in both the model and the TWR method, and it proved that the LMS Virtual.Lab Motion solver works really well.”
Leveraging upfront simulation
As a next step, the QRRS engineers employed the results from the multibody simulation to inspect NVH and the related durability behavior of the individual wagon components. By combining the calculated component-fatigue loads with material curves, cyclic fatigue parameters and stress results based on finite element analysis (FEA), LMS Virtual.Lab Durability can be used to accurately determine critical fatigue areas and assess the expected fatigue life.
“As all these solutions are integrated in one platform, this goes very smoothly,” says Wen Wang. “It is very straightforward to use results from one analysis as input for a next one. The software also has very effective manuals. Thanks to those and the expert support and training from the Siemens PLM Software staff, we could make very good progress with our analyses.” It helped QRRS engineers better understand the performance of their products: “By using the LMS solutions, we encountered some unforeseen problems early in design so we could take countermeasures to optimize our structure for durability. That has definitely improved the quality of our vehicle design,” says Wen Wang.
New methodologies trigger interest
The successful evaluation of the TWR technology for simulation also improved the collaboration between departments. The QRRS testing team, which uses LMS SCADASTM hardware and LMS Test.LabTM software for data acquisition and analysis, prepared the input data for the evaluation simulation in LMSTM Tecware software. This package helps engineers efficiently validate and understand gigabytes of raw mobile testing data.
The data signals are consolidated by various operations, such as removing anomalies, filtering and deriving new channels based on mathematical operations. In this way, the data can be prepared for further use in simulation. LMS Tecware allows subtracting durability-specific content and contains a wide range of dedicated data interpretation methods to help efficiently qualify and quantify the load data durability potential. The data can be readily imported into LMS Virtual.Lab Motion, the crucial component for connecting real-life testing to simulation for predicting fatigue life.
“During the evaluation, we had to work closely with our test colleagues,” says Wen Wang. “When they saw our final results, they were impressed. They would also like to include some of the capabilities of LMS Virtual.Lab Motion in their standard process. This is obviously a very good sign. It means that this new method gives us extra credibility inside the company and will help us to collaborate between departments. That will definitely help us increase quality while reducing time and cost when developing new vehicle structures in future.”
Transporting freight by train can be more energy efficient than by trucks or ship, especially over long distances. China annually transports over 4 billion tons of bulk cargo by rail, primarily coal, ores, minerals and grain. Driven by the enormous growth of its heavy industries, China plans to expand the network to 270,000 kilometers of track by 2050 to accommodate the expected increased demand for freight traffic capacity.
These rolling stock body structures undergo severe dynamic loading while in operation. The roughness of the rails, braking, curve passing and other cyclic or transient events subject them to enormous repeated stress over their lifetime. This can cause cracks and fractures in components such as the body, bogies, axles and bolsters. Therefore, avoiding fatigue damage is of primary importance when designing freight cars. On top of that, the market continuously demands faster transport in lighter vehicles to reduce time and cost. This could represent contradictory requirements, as higher speeds increase the dynamic loads, while reducing the amount of material could weaken the structure.
Railway engineers spend a substantial amount of development time trying to meet their customers’ financial considerations without compromising safety and durability. As so-called “design by experience” leads to recurrent prototype testing and long and expensive development cycles, QRRS experts are constantly looking for new durability engineering technologies, including simulation, that frontloads design decisions and delivers better performance on the first attempt, drastically shortening development time.
Efficient and reliable rail-load predictions
Crucial to successful fatigue life predictions is defining correct rail loads and understanding how they affect individual wagon components. Calculating them is challenging, especially because of the complex friction relation between the wheel and rails. Measurement data is usually either only partially available or not available at all during the concept phase.
Over the years, multibody simulation has been promoted as a reliable technology for accurately calculating rail loads. Several methodologies exist to handle different levels of model complexity and available measured data. Product lifecycle management (PLM) specialist Siemens PLM Software has developed a full computer-aided engineering (CAE) approach called “the digital test track,” as well as a combined test and CAE approach. The latter, included in LMS Virtual.Lab™ Motion software, is based on time waveform replication (TWR), originally a test technology for exact recreation of measured field data on a laboratory shaker system.
This process provides QRRS engineers with an efficient and reliable durability engineering approach. It gives them full insight into the dynamic behavior of all wagon components and helps them make the right decisions in the early stages of design.
“Noise, vibration and harshness (NVH), and especially its effect on durability, are critical design criteria for us,” says Wen Wang, team leader of simulation and analysis at QRRS. “So we want to consider them as early as possible in the vehicle development process. But at the concept stage of our design, there is obviously no prototype available and we simply cannot just get the loads from a rail test or a test rig.
“Simulation is the only option. But in the software packages we used, we had to create a digitized rail and needed to model the complex interaction between the rails and the wheel. This is very time consuming and difficult to do.
“The experts from Siemens PLM Software came up with a very pragmatic approach. The solution in LMS Virtual.Lab Motion allows us to start from measured displacements on an existing vehicle and calculate equivalent input signals for a new model. They can be applied to a fully detailed multibody model that simulates the dynamic response. These results include the loads on the individual components for NVH analysis or fatigue-life prediction, just like we wanted.”
Adopting the TWR approach
The TWR-based approach can start from any type and number of tested data. The process replicates a laboratory vehicle rail load test using an unconstrained multibody model and experimental data measured on a similar existing structure as boundary condition. Through an iterative control technique, back-calculated equivalent drive signals can be computed. By computing these drive signals for the wheel centers, LMS Virtual.Lab Motion doesn’t need to be used to model complex elements, e.g., friction models between the wheels and rails, and digitized rails.
Moreover, it eliminates the risk that the body model includes redundant constraints. The simulation corresponds to a durability test rig. LMS Virtual.Lab Motion features a powerful and accurate solver and dedicated modeling functionalities such as flexible bodies, as well as specific elements for fatigue analysis on spot and seam welds. The software then enables the user to calculate the individual component loads for use in a simulation with LMS Virtual.Lab Noise and Vibration software or LMS Virtual.Lab™ Durability software.
“Together with the experts from Siemens PLM Software, we validated this technology on one of our existing vehicles,” says Wen Wang. “Starting from a detailed multibody model and measured accelerations on wheel axle, body and certain points on the bolster, we calculated an equivalent displacement excitation that simulates a 12-channel virtual bench. That includes eight vertical and four horizontal channels for the front and rear bogie. We used this excitation for a forward dynamic response calculation.”
The results from LMS Virtual.Lab Motion corresponded well to the measurements in some observation points, both in time and frequency domains.
“We could see that the vertical and lateral response of the body and bolster were accurately simulated,” confirms Wen Wang. “The correlation coefficient between measured and calculated data approached one. And even though the model includes some complex elements like nonlinear bushes, springs and friction factors, it only took a limited number of iterations until we had convergence when calculating the loads. This gave us full confidence in both the model and the TWR method, and it proved that the LMS Virtual.Lab Motion solver works really well.”
Leveraging upfront simulation
As a next step, the QRRS engineers employed the results from the multibody simulation to inspect NVH and the related durability behavior of the individual wagon components. By combining the calculated component-fatigue loads with material curves, cyclic fatigue parameters and stress results based on finite element analysis (FEA), LMS Virtual.Lab Durability can be used to accurately determine critical fatigue areas and assess the expected fatigue life.
“As all these solutions are integrated in one platform, this goes very smoothly,” says Wen Wang. “It is very straightforward to use results from one analysis as input for a next one. The software also has very effective manuals. Thanks to those and the expert support and training from the Siemens PLM Software staff, we could make very good progress with our analyses.” It helped QRRS engineers better understand the performance of their products: “By using the LMS solutions, we encountered some unforeseen problems early in design so we could take countermeasures to optimize our structure for durability. That has definitely improved the quality of our vehicle design,” says Wen Wang.
New methodologies trigger interest
The successful evaluation of the TWR technology for simulation also improved the collaboration between departments. The QRRS testing team, which uses LMS SCADASTM hardware and LMS Test.LabTM software for data acquisition and analysis, prepared the input data for the evaluation simulation in LMSTM Tecware software. This package helps engineers efficiently validate and understand gigabytes of raw mobile testing data.
The data signals are consolidated by various operations, such as removing anomalies, filtering and deriving new channels based on mathematical operations. In this way, the data can be prepared for further use in simulation. LMS Tecware allows subtracting durability-specific content and contains a wide range of dedicated data interpretation methods to help efficiently qualify and quantify the load data durability potential. The data can be readily imported into LMS Virtual.Lab Motion, the crucial component for connecting real-life testing to simulation for predicting fatigue life.
“During the evaluation, we had to work closely with our test colleagues,” says Wen Wang. “When they saw our final results, they were impressed. They would also like to include some of the capabilities of LMS Virtual.Lab Motion in their standard process. This is obviously a very good sign. It means that this new method gives us extra credibility inside the company and will help us to collaborate between departments. That will definitely help us increase quality while reducing time and cost when developing new vehicle structures in future.”
