Incremental Gravity Application

A forum dedicated to users with questions regarding soil materials and elements.

forum currently locked

Moderator: Moderators

Locked
avytin
Posts: 75
Joined: Fri Dec 08, 2006 5:41 pm
Location: MiT

Incremental Gravity Application

Post by avytin »

Is there a way to apply the gravity incrementally in a model. e.g. lets say we have some quad and quadup elements with defined values of gravity, can we do the analysis in lets say 100 steps where g starts from 0.01g and increases to g?


I want this because I am using the Dafalias Manzari model and I want to make sure that the state parameters get updated correctly.

Thanks!
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Post by fmk »

put the gravity loads inside a LoadPattern with a Linear time series and then do a static analysis using load control. if you want the gravity built up using 100 steps do 100 analysis steps with a load control factor of 0.01.

after the gravity load analysis you can set the gravity loads to remain const using loadConst command for rest of analysis.
Lanka
Posts: 21
Joined: Mon Feb 19, 2007 6:38 pm
Location: Fugro

Post by Lanka »

Hi there,
I am trying to use OpenSees to do dynamic 2D effective analyses to evaulate the performance of a underground structure due to liquefaction induced slope failure.
I am using UCSD PDMY materials along with quadUP elements to model soil.

Due to the irregular ground surface (compose of peaks and valleys) and non-horizontal soil layers with significant stiffness contrast in some caseses, I'm having problem in initializing correct stress distribution before switch material stage to plastic. So I was wondeing Is there way to do incremental gravity application (for example in 100 steps) to initialize the stresses when we use quadUP elements along with PDMY material?

I define gravity load (i.e., $gra) in element defintion as follows:

element quadUP 24 4120 4119 4083 4085 100 6 $bulk $fmass $hPerm $vPerm 0 $gra 0

and I use following steps to define the gravity application

numberer RCM
system ProfileSPD
test NormDispIncr 1.0e-6 50 2
algorithm Newton
constraints Penalty 1.e18 1.e18
integrator Newmark 1.5 1.
analysis Transient

analyze 3 5.e3

Any help with this problem is appreciated.

Thanks!
avytin
Posts: 75
Joined: Fri Dec 08, 2006 5:41 pm
Location: MiT

Post by avytin »

how do you initialize your stress conditions? do you match elastic poissons ratio with K0 and then check to see if the stresses of the elastic step are correct?

If you are having problems with stiffness constrast you can choose any stiffness you want to get your elastic initial stresses right and then change the shear and bulk moduli to the right values by using the updateMaterials command. This might be the best option to my opinion.

I am not sure I can see any way to apply the load incrementally the way you define it. Maybe you can write a subroutine similar to the updatematerials command that can update the gravity in the quadup elements.
Lanka
Posts: 21
Joined: Mon Feb 19, 2007 6:38 pm
Location: Fugro

Post by Lanka »

Thanks for your suggestions.

When I was initializing the stress conditions, I did match elastic poissons ratios with Ko (i.e., poissons ratio=Ko/1+Ko). Now I tried with uniform stiffness every where, elastic initial stresses look reasonable and also I was able update material stage to elasto-plastic. After that I changed shear and bulk moduli to the exact vaules using updateMaterials command:

updateMaterials -material 1 shearModulus 1.332e7
updateMaterials -material 1 bulkModulus 6.658e7

Now I am getting following message

MatParameter::setDomain<Domain *theDomain> -NO RESULT

Is this an error message? Also, stresses and displacements are not changing when I solve with new properties. Is this a problem with updateMaterials command or Am I doing anything wrong?

Thanks!
avytin
Posts: 75
Joined: Fri Dec 08, 2006 5:41 pm
Location: MiT

Post by avytin »

which version of OS are you using?

if you go to the ucsd opensees site they have examples illustrating the use of the updatematerials command
(e.g. cyclic.ucsd.edu/opensees/soil2/soil2.txt)

if they do not work then there might be a problem with your version.

this command works fine with my version I have (latest version through cvs).
Lanka
Posts: 21
Joined: Mon Feb 19, 2007 6:38 pm
Location: Fugro

Post by Lanka »

I am using the latest version of OpemSees (2.1.0) which I downloaded from the website. Also I was following examples in the ucsd OpenSees site.......still I have problem. Is this version different from the vesrion through cvs?

Thanks!
Locked