Stress tensor and pressure relationship

  • Context: Graduate 
  • Thread starter Thread starter Seyn
  • Start date Start date
  • Tags Tags
    Fluid dynamic
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
10 replies · 1K views
Seyn
Messages
6
Reaction score
1
캡처.PNG


I'm studying with Currie's Fundamental Mechanics of Fluids, Fourth edition.
I'm confusing with the above statement; the force acting in the x2 direction is P2=s12n1 which is n1 direction.
Is there anybody help me to understand this subject?
 
Physics news on Phys.org
Seyn said:
the force acting in the x2 direction is P2=s12n1 which is n1 direction.
No, ##P_2=\sigma_{12}n_1## is the second component of a vector. That vector is the result of multiplying your original normal vector by a matrix, so can be in an arbitrary direction. Its second component is the component in the direction of your second basis vector.

Note that I think the author has tried to simplify the notation slightly, and I think he's done it in an unhelpful way. The point is that there's a normal vector ##\vec n## which he has chosen to be parallel to the first basis vector, so in component form it is ##(n_1,0,0)^T##. So I'd say that "the unit normal vector ... is ##n_1##" is possibly confusing the issue - ##n_1## is a component of a vector (in this case, the only non-zero component, but still only a component), not a vector itself.
 
Ibix said:
No, ##P_2=\sigma_{12}n_1## is the second component of a vector. That vector is the result of multiplying your original normal vector by a matrix, so can be in an arbitrary direction. Its second component is the component in the direction of your second basis vector.

Note that I think the author has tried to simplify the notation slightly, and I think he's done it in an unhelpful way. The point is that there's a normal vector ##\vec n## which he has chosen to be parallel to the first basis vector, so in component form it is ##(n_1,0,0)^T##. So I'd say that "the unit normal vector ... is ##n_1##" is possibly confusing the issue - ##n_1## is a component of a vector, not a vector itself.
Do you mean the point is,
##\vec{P}=(P_1,P_2,P_3)## and,
##P_j=\sigma_{ij}n_i##
so
##\vec{P}=\sigma_{ij}n_i \hat{e_j}##??
 
Ibix said:
Yes.
Thank you for your kind explanation!
 
Reply
  • Like
Likes   Reactions: Ibix
You are confused because the result the author gives is wrong. A unit vector in the 1 direction is (1,0,0). The components of the stress vector on the surface of constant x (i.e., 1) are ##P_1=\sigma_{ 11}##, ##P_2=\sigma_{ 12}##, and ##P_3=\sigma_{ 13}##. So the stress vector is ##(P_1,P_2,P_3)=(\sigma_{11},\sigma_{12},\sigma_{13})##
 
Chestermiller said:
You are confused because the result the author gives is wrong. A unit vector in the 1 direction is (1,0,0). The components of the stress vector on the surface of constant x (i.e., 1) are ##P_1=\sigma_{ 11}##, ##P_2=\sigma_{ 12}##, and ##P_3=\sigma_{ 13}##. So the stress vector is ##(P_1,P_2,P_3)=(\sigma_{11},\sigma_{12},\sigma_{13})##
Thanks for the reply. Do you mean the term ##n_i##s are the component of arbitrary surface vector##\vec{n}##, and the author chose an unit vector in the 1 direction as the surface vector for the first example which is ##(1,0,0)##? As a result, because ##n_1=1##, ##n_2=0##, ##n_3=0##, the surface force vector is ##P_j=\sigma_{ij} n_i=(1 \cdot \sigma_{11}+0 \cdot \sigma_{21}+0 \cdot \sigma_{31},1 \cdot \sigma_{12}+0 \cdot \sigma_{22}+0 \cdot \sigma_{23},1 \cdot \sigma_{13}+0 \cdot \sigma_{23}+0 \cdot \sigma_{33})=(\sigma_{11},\sigma_{12},\sigma_{13})##??
 
Seyn said:
Thanks for the reply. Do you mean the term ##n_i##s are the component of arbitrary surface vector##\vec{n}##, and the author chose an unit vector in the 1 direction as the surface vector for the first example which is ##(1,0,0)##? As a result, because ##n_1=1##, ##n_2=0##, ##n_3=0##, the surface force vector is ##P_j=\sigma_{ij} n_i=(1 \cdot \sigma_{11}+0 \cdot \sigma_{21}+0 \cdot \sigma_{31},1 \cdot \sigma_{12}+0 \cdot \sigma_{22}+0 \cdot \sigma_{23},1 \cdot \sigma_{13}+0 \cdot \sigma_{23}+0 \cdot \sigma_{33})=(\sigma_{11},\sigma_{12},\sigma_{13})##??
yes