Abstract
Purpose
A study of vibration attenuation of a vehicle occupant needs to start by integrating models of human body, cushion, seat suspension and vehicle. The latter needs to be carefully chosen to get the right balance between modelling complexity and prediction accuracy. This work considers two simplified vehicle models (quarter and half car) and studies how much the predictions of these models deviate from that of the full car model.
Methods
The seat suspension parameters were chosen to be those which minimized seat effective amplitude transmissibility. A genetic algorithm based optimization procedure was used for this exercise. A total of 28 model response parameters were investigated which included accelerations of different sections of the body and seat suspension, forces generated due to contact of the body with the seat and the backrest, forces and moments at the internal joints of the body and seat effective amplitude transmissibility.
Results
It was seen that when the input to the system was the random disturbances from an irregular road, the predictions of the quarter car model were distinctly different from those of the full car model while the half car model gave acceptable results in a few cases. However, when the input condition was changed to a bump on the road, the apparent superiority of the half car model was less prominent.
Conclusion
Thus, for a random road profile, it is better to choose the half car as a simplified vehicle model. For a bump road profile, the quarter car model, which is considerably simpler, maybe preferred if computing time or complexity is a constraint.
A study of vibration attenuation of a vehicle occupant needs to start by integrating models of human body, cushion, seat suspension and vehicle. The latter needs to be carefully chosen to get the right balance between modelling complexity and prediction accuracy. This work considers two simplified vehicle models (quarter and half car) and studies how much the predictions of these models deviate from that of the full car model.
Methods
The seat suspension parameters were chosen to be those which minimized seat effective amplitude transmissibility. A genetic algorithm based optimization procedure was used for this exercise. A total of 28 model response parameters were investigated which included accelerations of different sections of the body and seat suspension, forces generated due to contact of the body with the seat and the backrest, forces and moments at the internal joints of the body and seat effective amplitude transmissibility.
Results
It was seen that when the input to the system was the random disturbances from an irregular road, the predictions of the quarter car model were distinctly different from those of the full car model while the half car model gave acceptable results in a few cases. However, when the input condition was changed to a bump on the road, the apparent superiority of the half car model was less prominent.
Conclusion
Thus, for a random road profile, it is better to choose the half car as a simplified vehicle model. For a bump road profile, the quarter car model, which is considerably simpler, maybe preferred if computing time or complexity is a constraint.
| Original language | English |
|---|---|
| Pages (from-to) | 983-1001 |
| Number of pages | 19 |
| Journal | Journal of Vibration Engineering and Technologies |
| Volume | 9 |
| Early online date | 7 Feb 2021 |
| DOIs | |
| Publication status | Published - Jul 2021 |
Keywords
- Biodynamics
- Multibody dynamics
- Genetic algorithm
- Vehicle model
- Seat suspension
- Optimization