A simple question on M?

[Chan-Hee Park]

Hello,

I have a straightforward question on M in OGS5. Why is every boundary condition used for the benchmarks DISPLACEMENT? Neither Stress nor strain? 

[Wenqing Wang]

For the Dirichlet type BC,  the primary variable of the momentum balance equation,  DISPLACEMENT can be assigned.
For the Neumann type BC of the momentum balance equation, the traction on the boundary can be specified. In OGS5, we use  keywords  DISPLACEMENT_X1, DISPLACEMENT_Y1 or  DISPLACEMENT_Z1 to distinguish the traction in X, Y, or Z direction for  the Neumann type BC.

From the point of view of solid mechanics, strain can not be set  as boundary condition explicitly.

On 06/25/2018 08:06 AM, Chan-Hee Park wrote:

Hello,

I have a straightforward question on M in OGS5. Why is every boundary condition used for the benchmarks DISPLACEMENT? Neither Stress nor strain? 

--
You received this message because you are subscribed to the Google Groups "ogs-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ogs-users+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Chan-Hee Park]

OK. It is a matter of the primary variable that OGS5 solves. Good. 

1. Why was the primary variable in OGS5 not stress? 

2. How do we convert a certain stress condition for install 0.1MPa to an equivalent displacement in BC or IC? I guess we should integrate it in this case. 

I guess the 2nd question is an engineering question while the 1st one is a fundamental one. Please, WW.

PCH

[Wenqing Wang]

1. Why was the primary variable in OGS5 not stress?

>Only if we use the mixed FEM, we can set stress as one of the primary variables,
 

2. How do we convert a certain stress condition for install 0.1MPa to an equivalent displacement in BC or IC? I guess we should integrate it in this case. 

> It is not possible to convert stress condition as displacement BC without solving equation. On boundary, one can either measure displacement or load or both of them in experiment.

[Joshua Taron]

CH,

It is, mostly, terminology.

A neumann in displacement is traction. You are right though, most codes reduce to stress/strain to general "boundary" conditions. But they do that for clarity. And that is far easier and more understandable to work with. OGS5 took a more fundamental approach. I do not, necessarily, disagree with that decision. But I generally do.

With regard to stress as primary variable: Tradition and existence of formulations. But WW is right, as he tends to be,

Your request is complicated, as it tends to be.

Happy to chat. Betting WW is too.

Josh

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-users+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "ogs-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-users+unsubscribe@googlegroups.com.