Boundary Conditions in different direction

[YuanxinXiao]

Hi @GiudGiud

I want to impose a Dirichlet bc that follows the rod, but the direction of the rod is 45 degrees from the coordinate system x and y axes, so I multiply cos(45) in front of the disp_x of the FunctionDirichletBC, but I get the following error, is there any way can help me to realize it？

code:

[topx]
type = FunctionDirichletBC
variable = cos(45)* disp_x
boundary = 'frontbar_rightside_1'
function = '-1e-3sin(t)'
[]
[botx]
type = FunctionDirichletBC
variable = cos(45) * disp_x
boundary = 'frontbar_leftside_1'
function = '1e-3t'
[]

error:

expected '=' after parameter name '*', got 'd'�

[GiudGiud]

Hello

variable expects a name, not an expression so this variable = cos(45)* disp_x is not allowed.
What you have to do is set 4 BCs here:

• 1 for disp_x on frontbar_rightside_1 with cos(45) times the displacement (note that cos(45) might be wrong, best use sqrt(2)/2)
• 1 for disp_y on frontbar_rightside_1 with sin(45) times the displacement (note that sin(45) might be wrong, best use sqrt(2)/2
• same for the other boundary

[YuanxinXiao]

Thank you very much, as you said I applied x and y directions separately and multiplied their components in function and they worked well!

[topy]
type = FunctionDirichletBC
variable = disp_y
boundary = 'frontbar_rightside_1 '
function = '-2e-2sin(0.05pit)sqrt(2)/2'
[]
[boty]
type = FunctionDirichletBC
variable = disp_y
boundary = 'frontbar_leftside_1'
function = '2e-2sin(0.05pit)sqrt(2)/2'
[]
[topx]
type = FunctionDirichletBC
variable = disp_x
boundary = 'frontbar_rightside_1 '
function = '-2e-2sin(0.05pit)sqrt(2)/2'
[]
[botx]
type = FunctionDirichletBC
variable = disp_x
boundary = ' frontbar_leftside_1'
function = '2e-2sin(0.05pi*t)*sqrt(2)/2'
[]

But then I am facing a second problem, when I compress this rod on the ab-plane, the rod bends as shown in the picture, and I want to avoid this phenomenon. Before when the bar was parallel to the x-axis of the coordinate system, I always controlled disp_y = 0 for bc1 and bc2 to avoid this deformation, but now that the bar has a 45-degree pinch foot to the coordinate axis, I can't get the bar to contract freely in the direction of pressure by simply setting disp_y = 0 or disp_x = 0. Do you have a good method to make the bar not only straight but also free to contract?

[GiudGiud]

if you could use a dirichletBC to achieve what you want in the non-inclined case, you should be able to use this bc in the inclined case
https://mooseframework.inl.gov/source/bcs/PenaltyInclinedNoDisplacementBC.html