Dear Team,
Can we use RC joint Material with hinged beam also? I tried the same, but my lambda value in eigen analysis is coming negative. There is not softening in my joint material.
Regards,
Zeeshan Bhat
RC Joint Material
Re: RC Joint Material
Can you share your model?
-
Zeeshanbhat
- Posts: 153
- Joined: Wed May 27, 2020 3:48 pm
Re: RC Joint Material
Dear Team,
I modelled the same building again. Now lambda is coming positive, but there is one more problem. the time period in z direction is increasing drastically and my models modal participation mass occurs mostly only in z direction i.e. my model starts vibrating in Z direction in Ist mode.
Model Summary:
I want to create rigid offests ony in columns equal to depth of beam as per ASCE 41, but due to non availability of joint offset command in STKO. I created a rigid joint material with offset only in z direction. Without offset my modal analysis values are as given below. However after including the joint material its behaviour changes drastically. In practice there should be a little bit decrease in time period due to the rigid offset .
Kindly help me in this regard.
The model is attached below. Regards,
Zeeshan Bhat
I modelled the same building again. Now lambda is coming positive, but there is one more problem. the time period in z direction is increasing drastically and my models modal participation mass occurs mostly only in z direction i.e. my model starts vibrating in Z direction in Ist mode.
Model Summary:
I want to create rigid offests ony in columns equal to depth of beam as per ASCE 41, but due to non availability of joint offset command in STKO. I created a rigid joint material with offset only in z direction. Without offset my modal analysis values are as given below. However after including the joint material its behaviour changes drastically. In practice there should be a little bit decrease in time period due to the rigid offset .
Kindly help me in this regard.
The model is attached below. Regards,
Zeeshan Bhat
Re: RC Joint Material
Here's a working version of your model:
When you use the RCColumnJoint, STKO internally creates the end-offset using rigidLinks. the rigidLink is a multi-point (MP) constraint.
Unfortunately, this constraint will be attached to nodes that will be part of the rigidDiaphragm as well.
As you probably know, when using the Transformation constraint handler, a node cannot be part of 2 or more MP constraints. In fact, if you had looked at the modes of deformation, you would have seen that the floors and the columns are detached, because OpenSees cannot handle both rigidDiaphragm and rigidLink at the same nodes.
The solution is to use the Penalty constraint handler. However, the eigen command, internally uses the Transformation, unless you do a pre-analysis choosing the Penalty.
So what I did, is to add a "fake" analysis (0 steps and NO WIPE_ANALYSIS!!!) just before the eigen. In this way, it is used just as a set-up to choose the analysis components for the Eigen (specially the Penalty).
In this "fake" analysis it's very importat NOT TO USE the wipeAnalysis, otherwise the settings are gone!
Regarding how to choose proper values for the penalty parameters, here's a small script that you can use for frame problems.
The idea is to estimate the order of magnitude of the entries of your system matrix. It should not be precise, just the order of magnitude...
This is a small matlab script with the typical values used in your model. It turns out that you can use penalty values of 1.0e16 or 1.0e17, depending on the rounding errors:
Unfortunately, this constraint will be attached to nodes that will be part of the rigidDiaphragm as well.
As you probably know, when using the Transformation constraint handler, a node cannot be part of 2 or more MP constraints. In fact, if you had looked at the modes of deformation, you would have seen that the floors and the columns are detached, because OpenSees cannot handle both rigidDiaphragm and rigidLink at the same nodes.
The solution is to use the Penalty constraint handler. However, the eigen command, internally uses the Transformation, unless you do a pre-analysis choosing the Penalty.
So what I did, is to add a "fake" analysis (0 steps and NO WIPE_ANALYSIS!!!) just before the eigen. In this way, it is used just as a set-up to choose the analysis components for the Eigen (specially the Penalty).
In this "fake" analysis it's very importat NOT TO USE the wipeAnalysis, otherwise the settings are gone!
Regarding how to choose proper values for the penalty parameters, here's a small script that you can use for frame problems.
The idea is to estimate the order of magnitude of the entries of your system matrix. It should not be precise, just the order of magnitude...
This is a small matlab script with the typical values used in your model. It turns out that you can use penalty values of 1.0e16 or 1.0e17, depending on the rounding errors:
Code: Select all
E = 30000;
L = 3300;
b = 400;
h = 400;
A = b*h
I = b*h^3/12
EA = E*A
EI = E*I
kEA = EA/L
kEI = 4*EI/L
oom_axial = round(log10(kEA))
oom_flex = round(log10(kEI))
oom = round((oom_axial+oom_flex)/2)
power = oom+8
A = 160000
I = 2133333333.33333
EA = 4800000000
EI = 64000000000000
kEA = 1454545.45455
kEI = 77575757575.75757
oom_axial = 6
oom_flex = 11
oom = 9
power = 17-
Zeeshanbhat
- Posts: 153
- Joined: Wed May 27, 2020 3:48 pm
Re: RC Joint Material
Dear Team,
Thank you very much.
Zeeshan
Thank you very much.
Zeeshan