Linear Dashpots

View previous topic View next topic Go down

Linear Dashpots

Post  costas on Wed Jan 26, 2011 11:34 am

Does someone know: What does "velocity" stand for in DYNAPLOT command HYST>Viscous>Velocity
I expect this to be the relative velocity on the damper , i.e vel(i-end) - vel(j-end). However it doesn't.

In the case of SPRING this is valid, but in case of dampers it doesn't. Any idea?

costas

Posts : 22
Join date : 2010-01-12
Location : ROSE School, Pavia, Italy

Back to top Go down

Dashpots

Post  Athol Carr on Fri Jan 28, 2011 4:07 am

The Dynaplot options for the daspots are:
Damping forces, Velocities, displacements and stiffness factors(current C divided by initial C) The velocities are the velocity at the dashpot which is largely the difference in the nodal velocities but also, depending on the length of the dashpot member and the parameter SL (where in member length the dashpot acts) may be affected by the angular velocities at the nodes (similarly the rigid end links, if present will couple end rotational velocities to the velocity at the dashpot).
In Ruaumoko2d there are up to 3 forces, 3 velocities and 3 displacements and 3 stiffness factors. In Ruaumoko3D there are up to 6 of each.

Athol Carr

Posts : 115
Join date : 2009-12-09
Location : Christchurch, New Zealand

Back to top Go down

Re: Linear Dashpots

Post  costas on Fri Jan 28, 2011 1:35 pm

Thank you for the answer.
Even if I use zero-length dampers the difference of velocity between start- and end-node is not equal to the dashpot velocity plotted by DYNAPLOT>HYST>VISCOUS. The difference is very significant (perhaps 1/100). This seems weird...

costas

Posts : 22
Join date : 2010-01-12
Location : ROSE School, Pavia, Italy

Back to top Go down

Linear dashpots.

Post  Athol Carr on Fri Jan 28, 2011 8:30 pm

If the difference is of the order of 1% I would not be too concerned. I thought that you implied serious discrepancies. The data from Ruaumoko to Dynaplot is stored in single precision (32 bit), so there will be some loss in accuracy in that process. If you are then using Dynaplot to compute the differences in nodal displacements, you may be looking at small differences in large numbers. In Ruaumoko all calculation is in double precision (64bit) and so the element internal velocities are based on the differences in the more accurate numbers. In theory, one could store the numbers in the .RES file in 64bit form but that would double the size of the post-processor files and, in general, for post-processing that degree of precision is nor warranted.

Athol Carr

Posts : 115
Join date : 2009-12-09
Location : Christchurch, New Zealand

Back to top Go down

Re: Linear Dashpots

Post  costas on Fri Jan 28, 2011 9:24 pm

Sorry, I meant discrepancies of the order of 100 or even more....

costas

Posts : 22
Join date : 2010-01-12
Location : ROSE School, Pavia, Italy

Back to top Go down

Linear dashpots.

Post  Athol Carr on Fri Jan 28, 2011 9:58 pm

That sort if discrepancy is strange. No-one has ever noted problems with the dashpots over the years and enough have played with them. Maybe you need to send the data to illustrate the error you note so that it can be investigated..

Athol Carr

Posts : 115
Join date : 2009-12-09
Location : Christchurch, New Zealand

Back to top Go down

Re: Linear Dashpots

Post  costas on Sat Jan 29, 2011 9:22 am

Dr. Carr
I post some of my code so it can be viewed in case somebody has similar questions and I will send you the whole,
Model is a 30-node 29 m length pile. On each pile-node a zero-length spring and a dashpot are attached, and their other end is fixed (1,1,1). (Connection in parallel) The fixed end is excited with displacement TH's obtained from a site response analysis. Could the dashpots' response be influenced by the fact that they are connected in parallel with nonlinear springs?
FILE DATA:

Initial lines :
9 0 1 0 1 29 0 0 0 0 0
61 87 33 3 1 2 9.81 5 5 0.00001 1.00 1.0
2000 500 500 1 1 1 1 1.0
50 50

Definition of dampers:
4 DAMPER
0 2993.2 0 0 0 0 1 0 0 0 1 0 0

Let's focus on one damper:
61 4 4 35

The major problem is that:
DYNAPLOT>HYST>VISC elem 61 >X-VEL is completely different than
(DYNAPLOT>TIME>NODE>VEL n 4) minus (DYNAPLOT>TIME>NODE>VEL n 35)

costas

Posts : 22
Join date : 2010-01-12
Location : ROSE School, Pavia, Italy

Back to top Go down

Re: Linear Dashpots

Post  Sponsored content


Sponsored content


Back to top Go down

View previous topic View next topic Back to top


 
Permissions in this forum:
You cannot reply to topics in this forum