Constitutive Modelling Flashcards

1
Q

Linear-elastic vs Elasto-plastic

A

-L-E cannot simulate failure. Also assume E to be constant but it decreases with strain so elasto plastic normally applies

E-P:
Uniaxially loaded linearly elastic perfectly plastic bar
-L-E implies perfect recovery of strains
-PP implies no increase in stress with unloading elastic
* Constant E and unloading // to L-E curve
* Only partial recovery

Uniaxially loaded linearly elastic with strain hardening
- sigma exceeds sigma_y with a nonlinear increase
- Continuous increase up to ultimate failure
- Resembles an oedometer test

Uniaxially loaded linearly elastic with strain stiffening
- sigma falls under sigma_y with a nonlinear increase continuous decrease up to ultimate failure
- resembles shear box test for dense sand which softens as it tends to the residual strength

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Concepts for extension to general stress strain space

A

-Required since there are 6 stresses and strains rather than 1
-Concepts necessary
*Coincidence of axis
*Hardening/softening behaviour
*Yield function
*Plastic potential func.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Coincidence of axis

A

Direction of accumalated stress and incremental increase in strains, elastic strains in L-E and plastic ones after, coincide

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Yield function

A

F({sigma},{k}) = 0

  • No longer defined by one stress
  • F=0 is elasto-plastic , <0 is elastic and no state for >0 exists
  • k is the state parameter (Hardening softening rules)
  • Constant means perfectly plastic
  • Increasing means hardening
  • Decreasing means softening
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Plastic potential function

A

P({sigma},{m})=0

-Applied in the flow rule to get the direction of the plastic strain so m isnt important

delta epsilon_i_p = constant . partial dP/d.sigma_i
-epsilon_p is always perp. to the PP surface
-coincidence of axis allows sigma_i and delta epsilon_i_p to be plotted on the same exis
- Associated plasticity means PPF and YF re the same and the yield condition applies

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

2D rep. of yeild function
- hardening characteristics

A

Perfectly-plastic
- Yield function remains stationary
- No stress in y still leads to epsilon_y due to Poissons ratio for applied sigma_h
- Strain hardening
* Isotropic means yield function changes size
* Kinematic means the field function surface moves
* epsilon_p changes as the hardening happens

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Invariants

A

Principle stresses independent of axis unlike sigma_x, sigma_y, sigma_z
- Mean effective stress
-Deviator stress = 1/root(6) . root(sum(difference if the three stresses squared))
-Lodes angle = Tan inv(1/root(3) .
2(sigma2’ - sigma3’)/(sigma1’-sigma3’ - 1))

Significance:
- Mean effective stress is the distance from the deviator plane to the origin when scaled by root(3)
-Deviator stress is the distance form stress state P to the centre of the deviator plane when scaled by root(2)
-Lodes angle is the position of P measured clockwise form the horizontal
-Commonly used to interpret experimentals

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Space diagonal

A

All three stresses are equal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Deviatoric plane

A

plane perp. to the plane diagonal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Triaxial compression vs extension

A

Compression
-sigma1’ = sigma_a’
-sigma2’ = ‘sigma3’ = sigma_r’
-lodes angel = - 30 degrees

Extension
-sigma3’ = sigma_a’
-sigma1’ = ‘sigma2’ = sigma_r’
-lodes angel = 30 degrees

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Strain invariants

A

-Volumetric strain (epsilon_vol) = sum of the three
-Deviatoric strain (epsilon_d) = 2/root(6) . root(sum(difference of the strains squared))

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Trescas failure

A

F = sigma1 - sigma2 - 2Su = J.cos(theta) - Su = 0
- Yield surface is a regular hexagon on the deviatoric plane and since it independent o p’ is remains constant
-Amplitude of all stress states are Su

PPF:
-coincidence of axis allows J/epsilon_d and P/epsilon_vol plotted together
-the change in volumetric plastic strain is zero for associated plasticity
-Tresca is perfectly plastic

How well did you know this?
1
Not at all
2
3
4
5
Perfectly