The document discusses using simulation to find the optimal bucket filling strategy for a construction vehicle. It analyzed 5781 trajectories with 5 different filling strategies (labeled A through E) through discrete element method simulation. Various problems were encountered in the simulations, such as particles jumping unpredictably and large upward forces on the bucket, and solutions were developed to address them, such as adjusting the time step or adding more particle layers. Overall strategy A seemed most efficient while B was very inefficient. The analysis noted several important factors that were not captured in the simulations, such as the full loading cycle and stability of the vehicle, and suggested coupled large-scale simulation and optimization of the machine, environment, and operator may be needed to fully address finding the optimal strategy
20. …through simulationsimulation
Problem: “popcorn effect” of particles jumping around on their own account
(which does not change simulation results but looks funny)
Solution: adjust time step in the DEM simulation
(small penetrations are used to emulate microscopic deformations of particles)
(resistive, passive forces becoming driving, active ones)
DEM = Discrete Element Method NSCD = Non-Smooth Contact Dynamics
21. …through simulationsimulation
Problem: large upward force on bucket during initial penetration
(caused by inelastic particles getting stuck betw. cutting edge and ground)
Solution: add more layer of particles underneath the bucket
23. …through simulationsimulation
Problem: loss of material through landslide and spill
Solution: different trajectory that is more energy-efficient
- capture material from the landslide by moving forward?
- avoid spill by reversing more gradually
24. 5781 trajectories,
5 types
Towards finding
the optimal bucket filling strategy
through simulation
Fachtagung
Baumaschinentechnik
Dresden 2015
27. Towards finding
the optimal bucket filling strategy
through simulation
• ”A” seems like a good strategy to follow in general
• ”B” seems very inefficienct (no ”free penetration work”)
• ”C” requires more and finer steps (output in between A and B)
• ”D” requires an operator model in the loop
• ”E” requires better discretization (will reproduce A-D and beyond)
Disclaimer:
• ...
28. Towards finding
the optimal bucket filling strategy
through simulation
Disclaimer:
• It all has to be considered in a complete loading cycle (not only short loading cycle)
• Too great exit height of bucket from pile affects stability of the machine
• Overall efficiency is affected by the machine system (how power is generated and used)
• Other type of material (clay or shot rocks instead of gravel) will change the outcome
• Needed: large, coupled simulation/optimisation of machine, environment and operator
(perhaps soon possible thanks to cloud computation)