Adding damping values in a dynamic analysis

Question

I have a pile in a block of sand, with a horizontal load being applied at the top of the pile for 7 seconds, then being released allowing free vibration for 13 seconds. I added 1% damping to the pile and 0.6% damping to the soil materials. After this, the output computes no displacements in either stage of the analysis - loading or vibration. Without the damping values applied I get increased displacement until the load is deactivated, and then slowly decaying oscillations due to geometric dissipation. I expect to see the same output when I add the rayleigh damping values, except the oscillations should decay more. Why am I getting flat lines and how can I apply damping correctly?

Sean Johnson · Answer

It sounds like the damping values used may be dubious. 
 In PLAXIS FEM total damping is total of 4 independent processes: 1. The Rayleigh Damping 2. The material Damping (Modulus Reduction Curve) 3. Numerical Damping from Newmark parameter 4. Geometric Damping (ie. Waves spreading radially from the source having decreasing amplitude in proportion to the radius travelled). 
 The sum of these damping parts equals the total damping for the model. As a rule of thumb the second Rayleigh target frequency (f2) is typically 3 times the (f1). 
 You should take care not to overdamp your FEM model. The total damping of the system is the superposition of all the different components of damping. 
 
 A good approach may be to set the target frequencies as far apart as possible such that you cover as good as possible all relevant frequencies from both input frequencies and natural frequencies. You should try as good as possible to have a more or less constant damping over this range (since damping appears to be frequency independent). 
 
 Note that you can view the actual non-linear Rayleigh frequency - damping curve in the side panel of the material input window in Plaxis while defining alpha and beta.

Adding damping values in a dynamic analysis

Top Replies