Different Units for materials

[AdritaKundu]

Hi,

I have to create a FEM model of an RCC bridge pier retrofitted with CFRP wraps needed for my dissertation. I have used ConfinedConcrete01 for the core concrete which requires all the inputs to be defined in the MPa system. For simulating the Bond-Slip interaction at the interface of the column and the footing, I want to use the material BondSP01. But the problem is that BondSP01 operates only within the ksi system.

As there must be coherency in the units throughout the script(s), how do I overcome this difficulty? Is there any way to change the forces and displacements generated by the elements made from BondSP01 material in ksi system to MPa system, internally in the analysis?

Thanks in advance.
Adrita

[kesavapraba]

Dear Friend, First of all OpenSees does not have any option like SAP2000 or ETABS where you can easily change your units system. OpenSees allows the user to adopt the basic units system (N-mm-sec, kN-m-sec etc). Adopt anyone system as you wish, but be consistent with what units system throughout your model. I would suggest you to create a separate tcl file for units conversation if you want. For example , if suppose your basic unit system is in N, mm, sec and now you have force F of 100 kN. Then, you will have to do 'set F [expr 100*$kN]. But before doing it, you should have defined what that kN is in your units tcl file. That is your units file should have the line 'set kN [expr 1000.0]. There is already LibUnits.tcl file available in OS wiki page. Please refer that. All the best

[AdritaKundu]

kesavapraba wrote:
> Dear Friend, First of all OpenSees does not have any option like SAP2000 or
> ETABS where you can easily change your units system. OpenSees allows the
> user to adopt the basic units system (N-mm-sec, kN-m-sec etc). Adopt anyone
> system as you wish, but be consistent with what units system throughout
> your model. I would suggest you to create a separate tcl file for units
> conversation if you want. For example , if suppose your basic unit system
> is in N, mm, sec and now you have force F of 100 kN. Then, you will have to
> do 'set F [expr 100*$kN]. But before doing it, you should have defined what
> that kN is in your units tcl file. That is your units file should have the
> line 'set kN [expr 1000.0]. There is already LibUnits.tcl file available in
> OS wiki page. Please refer that. All the best

Hello,

I completely understand what you mean. And if I were to work with the standard uniaxial materials in OpenSees, I would have taken these very steps mentioned. However, I am using these contributed materials (ConfinedConcrete 01 and BondSP01). There's a script in the source code of BondSP01 which mentions explicitly that the input must be in psi system which means that I can't use MPa system (even after proper conversion) to define any material whatsoever for this material type as it would result in absurd forces and deformations. Also, as per the wiki page for ConfinedConcrete01 uniaxial material, it was mentioned in an important note that "the inputs to be used are in MPa, mm.".

This is where I am facing my problem. Both the materials can only be used in specific unit systems which happen to be different from each other. This is leading to incoherency in units and hence, absurd results. Can you please help me in this aspect.

Thanks in advance.
Adrita

[kesavapraba]

Dear Friend, Why don't you try creating two different system of units tcl files and source the file at just before appropriate places? For example, you have units system in N.mm.sec at start of your code, then if you want to define BondSlip material, you source psi based units tcl file. Once that definition is over, immediately below that material command again source N.mm.sec based units tcl file. Just try it. Thanks

[AdritaKundu]

kesavapraba wrote:
> Dear Friend, Why don't you try creating two different system of units tcl
> files and source the file at just before appropriate places? For example,
> you have units system in N.mm.sec at start of your code, then if you want
> to define BondSlip material, you source psi based units tcl file. Once that
> definition is over, immediately below that material command again source
> N.mm.sec based units tcl file. Just try it. Thanks

Hello,

Have you ever applied this solution to such a similar problem? Because I am afraid that what you are suggesting will not work. The system of units imposed must be consistent throughout the program. If not, then the program will lead to absurd forces and deformations.

I will try your suggestion and will let you know about the results.

Thanks.

[AdritaKundu]

Hi Kesav,

I tried it and as expected, it failed to work.

I will now change the source code for this material to convert its units system to MPa system. I don't know how to implement the module into OpenSees. Can you please help me regarding that.

Thanks in advance.
Adrita

[kesavapraba]

Dear Friend, Have tried as per the protocol I mentioned ? See, say there are two units tcl files, then you are calling one file at start of your code. Now comes the definition of your BondSlip material which requires the units in psi..So just above the line where you define the material, you call the 2nd units tcl file which has the psi as base unit. This is where the whole conversation takes place. Now you can give material parameters in MPa for BondSlip material. Once done wthe definition, immediately in the below line, you again call the default units tcl file.