What does "traction free" mean? Is it a boundary condiction that there's no stress action on that boundary? For example, the pseudio-fluid is free to leave the mold wall as pushed out by the advancing liquid.

Thanks a lot!

Traction-free = no surface traction = no tangential stress component.

Thanks for the reply! The boundary condition is, sigma(u).n-p.n=0 in the paper I read. I don't know the mean. no tangential stress component and no normal stress? Thanks a lot!

generally traction is used for stress bc in literature.

>sigma(u).n-p.n=0

n probably mean normal vector to boundary and .n mean normal component of those vector, as p (probably pressure) is scaler the better expression is pI.n

Strictly the zero traction condition is

(Tn).s = 0

where T is the stress tensor, n is the normal to the surface and s is the surface tangent (in 3D there are two surface tangents). Note that, since n.s =0 the pressure does not contribute to this calculation and u.n=0 for no flow through the surface. For n=k ans s=i then this is simplifies to the usual conditions (laminar Newtonian fluid)

w=0, u_z=0.

Note that traction free surface mean no stresses normal stress and tangent one,

the general stress bc condition for an arbitrary 3d surface is (in fluid dynamics):

i. (S.n).n = normal stress e.g. surface tension = sigma k/We

S= -pI + mu/Re[(grad U)+ (grad U)^T] (stress tensor)

where n is local normal vector, mu is viscosity, p is pressure, U is velosity vector, k is kurvature, Re is renold number, We is weber number

ii. (S.n).m1=0 and (S.n).m2=0

where m1 and m2 are local tangential vectors, note that n, m1 and m2 make independent basis for 3d space

Finally traction free mean condition leads to:

(S.n).n =0. (S.n).m1 =0. (S.n).m2 =0.

Is it clear yang?

The condition (S.n).n =0 will over prescibe the system if you require u.n=0 on a solid surface. The only time you would employ a condition on (S.n).n is if you had a surface which could deform under the motion of the fluid (even in this case it's really a condition on p rather than on S.n).n)).

i don't knowe what is your argue about correctness of previous massage, (S.n).n =0 in solid mechanics this condition is usually natural on free boundaries of solid but in fluid dynamics, normal component usually balance with pressure force (ambient pressure+surface tension pressure) i.e.

(S.n).n = P_ambient + sigma k

is it correct?

My point is that if you have a "traction free" or freeslip rigid surface then u.n = 0 is the relevant boundary condition and it does not in general imply (Sn).n=0; i.e. there is a normal pressure force on the surface even if there is no traction (think inviscid flow past an aerofoil - there is no drag but there is lift).

well, you mean traction free as only tangent (or shear) stresse=0., althouth i don't see this expression in CFD literature but in solid mechanics it is general based my nowledge (normal+tangent components)

Thanks for your replies! The air treated as pseudio-fluid is free to leave the mold wall as pushed out by the advancing liquid.The traction free boundary condition is T.n=0 at the interface between the wall and the air. Where, T=-pI+2*mu*D,D=0.5*[grad(u)+grad(u)^T]

In fluid mechanics the normal condition is generally only applied to the pressure (on a material surface); i.e. constant atmospheric pressure or a pressure jump related to surface tension - this is why it is usually referred to as a stress-free surface and not traction free one.

[psb]

Hi..

I want to implement the condition –p+0.0035 (∂u/∂x)=0 stated as the traction free boundary condition at the outlet of a pipe. How to implement this in Fluent? Is there any such option in Fluent or it will require a udf?

[guanjiang.chen]

Originally Posted by Tom
;44323

Strictly the zero traction condition is

(Tn).s = 0

where T is the stress tensor, n is the normal to the surface and s is the surface tangent (in 3D there are two surface tangents). Note that, since n.s =0 the pressure does not contribute to this calculation and u.n=0 for no flow through the surface. For n=k ans s=i then this is simplifies to the usual conditions (laminar Newtonian fluid)

w=0, u_z=0.

This seems to be slip boundary condition.