yes, i believe it is calibrated as such.
i have never used it, yet.
About Bond_SP01 Command
Moderators: silvia, selimgunay, Moderators
-
xiaoqg2003
- Posts: 10
- Joined: Tue Jul 24, 2007 5:06 am
- Location: zhejiang Province, PR China
Dear Dr. Silvia,
Also, I found that the bondsp01 model only slightly change the initial stiffness of the specimens. This is reasonable, as at the initial loading stage, the slip between the concrete and reinforcing steel is small.
That is another problem. If the initial stiffness obtained from test results is much different from that of simulation, which means additional rotation exists besides bond slip, do i need also to use spring element (in another zerolength element) to fit the initial stiffness. Is there any better method? For example, can I consider the spring and bond slip in just one zerolength element while in the previous consideration, I need to conbime two zerolength elements in series connection.
Thanks
Also, I found that the bondsp01 model only slightly change the initial stiffness of the specimens. This is reasonable, as at the initial loading stage, the slip between the concrete and reinforcing steel is small.
That is another problem. If the initial stiffness obtained from test results is much different from that of simulation, which means additional rotation exists besides bond slip, do i need also to use spring element (in another zerolength element) to fit the initial stiffness. Is there any better method? For example, can I consider the spring and bond slip in just one zerolength element while in the previous consideration, I need to conbime two zerolength elements in series connection.
Thanks
-
xiaoqg2003
- Posts: 10
- Joined: Tue Jul 24, 2007 5:06 am
- Location: zhejiang Province, PR China
I found if I used the equation recommended by Dr. Zhao (the presenter of bondsp01). The initial tangential stiffness of bondsp01 (about 3000) is approximately 10 percent of that of the corresponding steel (about 29000). The initial stiffness is generally reduced no more than 7% as the tangential stiffness of concrete used in the zerolength element can not be changed.
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).
Hi, I am very curious on your comments. Did you already have some experimental work showing that the strain penetration influence the initial stiffness? The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well.
[quote="xiaoqg2003"]I found if I used the equation recommended by Dr. Zhao (the presenter of bondsp01). The initial tangential stiffness of bondsp01 (about 3000) is approximately 10 percent of that of the corresponding steel (about 29000). The initial stiffness is generally reduced no more than 7% as the tangential stiffness of concrete used in the zerolength element can not be changed.
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).[/quote]
[quote="xiaoqg2003"]I found if I used the equation recommended by Dr. Zhao (the presenter of bondsp01). The initial tangential stiffness of bondsp01 (about 3000) is approximately 10 percent of that of the corresponding steel (about 29000). The initial stiffness is generally reduced no more than 7% as the tangential stiffness of concrete used in the zerolength element can not be changed.
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).[/quote]
Yuhao Feng
Research Assistant & graduate student
CCEE Dept. North Carolina State University
Research Assistant & graduate student
CCEE Dept. North Carolina State University
-
xiaoqg2003
- Posts: 10
- Joined: Tue Jul 24, 2007 5:06 am
- Location: zhejiang Province, PR China
I think that my points are as follows:
From simulation results, when we use bond_sp01 model, it actually can reduce the inital stiffness (or the secant modulus of the first cycle). My understanding is that the bond_sp01 can be seen as a weak material compared to steel.
In some experimental case, the initial stiffness of the test result is still smaller than prediction even bondsp01 is used.
At last, I agree your comment that conceptually bond_sp model should not influence the initial stiffness, as the strain penetration should not occur in the preliminary stage.
about your point "The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well."
I think for a simulation if you can choose proper test data, you can always match well at most of the time
[quote="yfeng"]Hi, I am very curious on your comments. Did you already have some experimental work showing that the strain penetration influence the initial stiffness? The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well.
[quote="xiaoqg2003"]I found if I used the equation recommended by Dr. Zhao (the presenter of bondsp01). The initial tangential stiffness of bondsp01 (about 3000) is approximately 10 percent of that of the corresponding steel (about 29000). The initial stiffness is generally reduced no more than 7% as the tangential stiffness of concrete used in the zerolength element can not be changed.
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).[/quote][/quote]
From simulation results, when we use bond_sp01 model, it actually can reduce the inital stiffness (or the secant modulus of the first cycle). My understanding is that the bond_sp01 can be seen as a weak material compared to steel.
In some experimental case, the initial stiffness of the test result is still smaller than prediction even bondsp01 is used.
At last, I agree your comment that conceptually bond_sp model should not influence the initial stiffness, as the strain penetration should not occur in the preliminary stage.
about your point "The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well."
I think for a simulation if you can choose proper test data, you can always match well at most of the time
[quote="yfeng"]Hi, I am very curious on your comments. Did you already have some experimental work showing that the strain penetration influence the initial stiffness? The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well.
[quote="xiaoqg2003"]I found if I used the equation recommended by Dr. Zhao (the presenter of bondsp01). The initial tangential stiffness of bondsp01 (about 3000) is approximately 10 percent of that of the corresponding steel (about 29000). The initial stiffness is generally reduced no more than 7% as the tangential stiffness of concrete used in the zerolength element can not be changed.
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).[/quote][/quote]
I am not sure. My suggestion is to talk to the author. I believe he can give the answer you want.
[quote="xiaoqg2003"]I think that my points are as follows:
From simulation results, when we use bond_sp01 model, it actually can reduce the inital stiffness (or the secant modulus of the first cycle). My understanding is that the bond_sp01 can be seen as a weak material compared to steel.
In some experimental case, the initial stiffness of the test result is still smaller than prediction even bondsp01 is used.
At last, I agree your comment that conceptually bond_sp model should not influence the initial stiffness, as the strain penetration should not occur in the preliminary stage.
about your point "The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well."
I think for a simulation if you can choose proper test data, you can always match well at most of the time
[quote="yfeng"]Hi, I am very curious on your comments. Did you already have some experimental work showing that the strain penetration influence the initial stiffness? The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well.
[quote="xiaoqg2003"]I found if I used the equation recommended by Dr. Zhao (the presenter of bondsp01). The initial tangential stiffness of bondsp01 (about 3000) is approximately 10 percent of that of the corresponding steel (about 29000). The initial stiffness is generally reduced no more than 7% as the tangential stiffness of concrete used in the zerolength element can not be changed.
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).[/quote][/quote][/quote]
[quote="xiaoqg2003"]I think that my points are as follows:
From simulation results, when we use bond_sp01 model, it actually can reduce the inital stiffness (or the secant modulus of the first cycle). My understanding is that the bond_sp01 can be seen as a weak material compared to steel.
In some experimental case, the initial stiffness of the test result is still smaller than prediction even bondsp01 is used.
At last, I agree your comment that conceptually bond_sp model should not influence the initial stiffness, as the strain penetration should not occur in the preliminary stage.
about your point "The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well."
I think for a simulation if you can choose proper test data, you can always match well at most of the time
[quote="yfeng"]Hi, I am very curious on your comments. Did you already have some experimental work showing that the strain penetration influence the initial stiffness? The reason I am asking this is that they generated the Bondsp01 model and compared it with test results also which match so well.
[quote="xiaoqg2003"]I found if I used the equation recommended by Dr. Zhao (the presenter of bondsp01). The initial tangential stiffness of bondsp01 (about 3000) is approximately 10 percent of that of the corresponding steel (about 29000). The initial stiffness is generally reduced no more than 7% as the tangential stiffness of concrete used in the zerolength element can not be changed.
That is why I think sometimes using a spring element to fit the initial sthiffness is necessary.
Actually, in Dr. Zhao's example, it seems that he also useed to beam elements to fit the initial stiffness (considering the stiffness of column fotting).[/quote][/quote][/quote]
Yuhao Feng
Research Assistant & graduate student
CCEE Dept. North Carolina State University
Research Assistant & graduate student
CCEE Dept. North Carolina State University
-
DanieleBaraldi84
- Posts: 5
- Joined: Thu Dec 18, 2008 4:02 am
- Location: Università degli Studi di Ferrara
implementation of Bond_SP01
Hi everybody,
I put my message here only to don't create another topic. My question is about the implementation of this material in OpenSees.
In this page..
www.uwm.edu/~jzhao/Bond_SP01_pages/Bond_index.html
..there is written by the material's authors that they have fixed a mistake in the original code, and they put a link for downloading the new one. Looking on the source code directories it seems that the current version of OpenSees still has the old bond-slip model (dated 2006).
For now with my small experience with OpenSees I had been only able to create a new material modifying the ElasticPP example made in c++ and using it by copying/pasting the .dll file into the work directory.
My question is if it's possible to replace an existing material by copying/pasting a .dll file with the same name or if I want to try the new suggested material I have to create a new one.
Thanks very much.
I put my message here only to don't create another topic. My question is about the implementation of this material in OpenSees.
In this page..
www.uwm.edu/~jzhao/Bond_SP01_pages/Bond_index.html
..there is written by the material's authors that they have fixed a mistake in the original code, and they put a link for downloading the new one. Looking on the source code directories it seems that the current version of OpenSees still has the old bond-slip model (dated 2006).
For now with my small experience with OpenSees I had been only able to create a new material modifying the ElasticPP example made in c++ and using it by copying/pasting the .dll file into the work directory.
My question is if it's possible to replace an existing material by copying/pasting a .dll file with the same name or if I want to try the new suggested material I have to create a new one.
Thanks very much.
Daniele Baraldi
I have communicated to twenty Bond_SP01 users. I would like to say a few things here.
1) Please do send me emails (jzhao@uwm.edu) if you found bugs. For that I would like the Opensees to be able record stress-strain data of fibers in zerolength section element for you to check your model input. You might want to go to the Bond_SP01 webpage and check out my answers to some frequently asked questions (this includes the use of EqualDOF).
2) The model formulation is unitless. There have been several applications of the model in SI units. I would ignore the warning. Again please visit the Bond_SP01 webpage for the related documents.
3) The model at the current version is for modeling the strain penetration behavior of steel rebars at member intersections. However please note that the envelope curve of the stress-slip relationship is only for fully anchored rebars. Please refer to the ACI paper for our thoughts on "fully anchored rebars." This limitation is due to fact that there are only two sets of load-slip test data available in the literature (to my knowledge). Please share your data for the further development of the model.
4) Continue on item 3), fully anchored bars are typically seen in columns, and the bars are embedded in foundations and bridge cap beams with straight ends or hooked ends. One needs to be careful when he/she would like to use the zerolength section element with Bond_SP01 for beam-column connections. In my example of bridge column connected to a cap beam, the model does represent the structural behavior, and the inclusion of the additional elastic element was backed by experimental data.
5) continue on Item 4), the bond-slip behavior of beam bars in beam-column joints is very likely different from those of column bars in foundations. The current envelope curve should be used with caution for beam bars in joints. More research is needed, including the bond-slip behaviors at roof corner joints.
My last comment is that understanding the structural behavior and having expected/observed behavior should always come before any manipulation of model parameters.
1) Please do send me emails (jzhao@uwm.edu) if you found bugs. For that I would like the Opensees to be able record stress-strain data of fibers in zerolength section element for you to check your model input. You might want to go to the Bond_SP01 webpage and check out my answers to some frequently asked questions (this includes the use of EqualDOF).
2) The model formulation is unitless. There have been several applications of the model in SI units. I would ignore the warning. Again please visit the Bond_SP01 webpage for the related documents.
3) The model at the current version is for modeling the strain penetration behavior of steel rebars at member intersections. However please note that the envelope curve of the stress-slip relationship is only for fully anchored rebars. Please refer to the ACI paper for our thoughts on "fully anchored rebars." This limitation is due to fact that there are only two sets of load-slip test data available in the literature (to my knowledge). Please share your data for the further development of the model.
4) Continue on item 3), fully anchored bars are typically seen in columns, and the bars are embedded in foundations and bridge cap beams with straight ends or hooked ends. One needs to be careful when he/she would like to use the zerolength section element with Bond_SP01 for beam-column connections. In my example of bridge column connected to a cap beam, the model does represent the structural behavior, and the inclusion of the additional elastic element was backed by experimental data.
5) continue on Item 4), the bond-slip behavior of beam bars in beam-column joints is very likely different from those of column bars in foundations. The current envelope curve should be used with caution for beam bars in joints. More research is needed, including the bond-slip behaviors at roof corner joints.
My last comment is that understanding the structural behavior and having expected/observed behavior should always come before any manipulation of model parameters.