E.G.J. Groen
Modeling the dynamic behavior of cable and belt systems in ADAMS
Computer program,
Report 2001.TT.5546, Transport Engineering and Logistics.
The Material Handling and Transport Engineering section has set up a project
called "Dynamic Modeling of Flexible Belt Systems" aiming at
developing models to simulate the dynamic behavior of cable systems and belt
conveyors. At present within the section a toolbox is being developed in the
multi body software package ADAMS to model cable systems according to the
massless cable method.
The primary research objective is to verify and test the toolbox developed in
the section. Verification implies the creation of models (test cases) and the
comparison the simulation results with other results from:
- simulation models according to other methods (for example the discrete
cable method).
- analytical solutions.
- empirical data from experiments on cable and belt systems.
Characteristic for the toolbox is the massless belt. The belt mass is
represented as an additional moment of inertia added to the pulleys. The
pulleys represent all the masses and motion of the system. The belt is
reduced to a spring-damper force. Because the belt cannot deform in
transverse direction the simulation of transverse vibration and sag is
tricked by adding extra pulleys, which can make free translational motions
in the transverse direction.
Two different test cases are carried out:
- The objective of the first case is to verify the calculation of belt
forces, strain, stiffness and natural frequencies. The model consists
of a fixed pulley and a hanging pulley. The cable force, strain and the
systems natural frequencies are measured and compared to analytical
solved values.
From the results it can be concluded that the results are accurate enough
(<3%) only for cable loads F < 0.003 * EA. As a result
of an incorrect method for the calculation of the cable stiffness the
results are inaccurate for larger cable loads. The calculation method can
be improved by using the untensioned cable part length instead of the total
cable part length.
- The objective of the second case is to verify the simulation of transverse
vibration of a stationary moving flat belt span supported by two idlers.
The simulation results are compared to experimental data and analytical
approximations. In this test case the model, described in the thesis of
Mr. Lodewijks (1996, chapter 9), is used. In total fifteen simulations
are performed with five different parameter sets. Most of the results are
reasonably accurate. The method for simulating the transverse vibration
with transverse moving pulleys works. The accuracy of the transverse
vibration increases with the number of extra pulleys used. For cases with
a moving belt the results are becoming worse when the model accuracy is
increased. From the results it seems that the model is not accurate enough
to simulate moving belts.
Reports on Transport Engineering and Logistics (in Dutch)
Modified: 2006.12.28;
logistics@3mE.tudelft.nl
, TU Delft
/ 3mE
/ TT
/ LT.