• Home
  • Textbooks
  • A Course on Plasticity Theory
  • Energy, stress, dissipation, and plastic evolution

A Course on Plasticity Theory

David J. Steigmann

Chapter 6

Energy, stress, dissipation, and plastic evolution - all with Video Answers

Educators


Chapter Questions

Problem 1

Assume $\mathbf{A}$ and $\mathbf{B}$ to be invertible and show that $(\mathbf{A B})^*=\mathbf{A}^* \mathbf{B}^*$.

Check back soon!
01:01

Problem 2

Prove the identity $\mathbf{A} \cdot \mathbf{B C D}=\mathbf{B}^t \mathbf{A D}^t \cdot \mathbf{C}$.

Raj Bala
Raj Bala
Numerade Educator
01:19

Problem 3

The classical stress-strain relation for isotropic materials is
$$
\hat{\mathbf{S}}=\lambda(t r \hat{\mathbf{E}}) \mathbf{I}+2 \mu \hat{\mathbf{E}},
$$
where $\lambda$ and $\mu$, the Lamé moduli, are constants (see the Appendix to this section). Verify (6.28) for arbitrary orthogonal $\mathbf{R}$.

Dominador Tan
Dominador Tan
Numerade Educator
13:37

Problem 4

Prove this.

Mahnoor Amin
Mahnoor Amin
Numerade Educator

Problem 5

If $\Omega$ is a material surface, then the referential and spatial surface dislocation densities are not independent. Derive the relationship between them, and show that if one vanishes, then both vanish.

Check back soon!
04:27

Problem 6

Show that $\operatorname{Curl} \mathbf{F}$ vanishes in the regions separated by the surface $\Omega$, assuming $\chi$ to be a twice-differentiable function of $\mathbf{x}$ therein.

Chai Santi
Chai Santi
Numerade Educator
03:16

Problem 7

Show that the dissipation D is invariant under material symmetry transformations.

Chai Santi
Chai Santi
Numerade Educator
07:02

Problem 8

Carry out the details.

Khoobchandra Agrawal
Khoobchandra Agrawal
Numerade Educator

Problem 9

Prove this.

Check back soon!

Problem 10

Using (7.54), show that $\mathbf{T}=T_{\alpha \beta} \beta^{\mathbf{i} \alpha} \otimes \mathbf{i}_\beta-p \mathbf{i}_3 \otimes \mathbf{i}_3$, where
$$
T_{11}=-p-k \sin 2 \theta, \quad T_{22}=-p+k \sin 2 \theta, \quad \text { and } \quad T_{12}=T_{21}=k \cos 2 \theta .
$$
Equilibrium without body force, i.e., $T_{i j j}=0$, thus requires that $p_{, 3}=0$ and $T_{\alpha \beta, \beta}=0$, the latter yielding
$$
p_{, 1}+2 k\left(\theta_{, 1} \cos 2 \theta+\theta_{, 2} \sin 2 \theta\right)=0 \text { and } \quad p_{, 2}+2 k\left(\theta_{, 1} \sin 2 \theta-\theta_{, 2} \cos 2 \theta\right)=0
$$
Note that if, at a particular point, we orient the basis $\left\{\mathbf{i}_\alpha\right\}$ such that $\mathbf{i}_1=\mathbf{t}$ and $\mathbf{i}_2=\mathbf{s}$, then $\theta$ vanishes at that point, and these equations reduce to
$$
\begin{aligned}
& 0=p_1+2 k \theta_{, 1}=\mathbf{i}_1 \cdot \operatorname{grad}(p+2 k \theta)=\mathbf{t} \cdot \operatorname{grad}(p+2 k \theta), \quad \text { and, similarly, } \\
& 0=p_{12}-2 k \theta_{, 2}=\mathbf{i}_2 \cdot \operatorname{grad}(p-2 k \theta)=\mathbf{s} \cdot \operatorname{grad}(p-2 k \theta),
\end{aligned}
$$
which, of course, are just the Prandtl-Hencky equations.
With the Cartesian form of the equations in hand, the obvious thing to do when searching for simple solutions is to assume that $p$ or $\theta$ depends on just one of the coordinates. Let's consider the possibility that $\theta$ depends only on $y_2$; Eqs. (7.66) reduce to
$$
p_{, 1}=k(\cos 2 \theta)^{\prime} \quad \text { and } \quad p_{, 2}=k(\sin 2 \theta)^{\prime}
$$
where $(\cdot)^{\prime}=d(\cdot) / d y_2$. Assuming $p$ to be twice differentiable, we can eliminate it as follows:
$$
k(\cos 2 \theta)^{\prime \prime}=p, 12=p_{, 21}=k(\sin 2 \theta)_{, 21}=0 .
$$
Then $\cos 2 \theta$ is a linear function of $y_2$. One such solution, due to Prandtl, is
$$
\cos 2 \theta=-y_2 / h, \quad \sin 2 \theta=-\sqrt{1-\left(y_2 / h\right)^2},
$$
where $h$ is a positive constant. This is meaningful provided that $\left|y_2\right| \leq h$, and from (7.68) we have
$$
p=-k\left(y_1 / h\right)+f\left(y_2\right), \text { where } f^{\prime}=k(\sin 2 \theta)^{\prime} .
$$
Thus,
$$
p=p_0-k\left[y_1 / h+\sqrt{1-\left(y_2 / h\right)^2}\right]
$$
where $p_0$ is a constant, and the equilibrium stress field follows from the solution to Problem 7.10. Our solution is adapted from Prager's version of Prandtl's solution.
To obtain the velocity field, we invoke (7.25) with
$$
\tau=k \sin 2 \theta\left(\mathbf{i}_2 \otimes \mathbf{i}_2-\mathbf{i}_1 \otimes \mathbf{i}_1\right)+k \cos 2 \theta\left(\mathbf{i}_1 \otimes \mathbf{i}_2+\mathbf{i}_2 \otimes \mathbf{i}_1\right)
$$
Then,
$$
D=\alpha\left(i_1 \otimes i_1-i_2 \otimes i_2\right)+\beta\left(\mathbf{i}_1 \otimes \mathbf{i}_2+\mathbf{i}_2 \otimes \mathbf{i}_1\right)
$$
where
$$
\alpha=-\lambda k \sin 2 \theta \text { and } \beta=\lambda k \cos 2 \theta
$$
Because $\lambda \geq 0$ we then have $\lambda k=\sqrt{\alpha^2+\beta^2}$ and
$$
\alpha / \sqrt{\alpha^2+\beta^2}=-\sin 2 \theta, \quad \beta / \sqrt{\alpha^2+\beta^2}=\cos 2 \theta,
$$
Dxford University Press (2023).pdf
Open with
$$
\alpha=\sqrt{1-\left(y_2 / h\right)^2} \sqrt{\alpha^2+\beta^2} \text { and } \beta=-\left(y_2 / h\right) \sqrt{\alpha^2+\beta^2} \text {. }
$$
It follows from the expression for $\mathbf{D}$ that the components of the velocity field satisfy
$$
v_{1,1}=-v_{2,2}=\alpha \text { and } \quad \frac{1}{2}\left(v_{1,2}+v_{2,1}\right)=\beta .
$$
Accordingly, $\alpha$ and $\beta$ are not arbitrary. Assuming the velocity components to be thrice differentiable, we derive the compatibility condition
$$
\alpha_{, 11}-\alpha_{, 22}+2 \beta_{, 12}=0
$$
which, with reference to $(7.77)$, is easily seen to be solved if
$$
\sqrt{\alpha^2+\beta^2}=\frac{C}{h}\left(\sqrt{1-\left(y_2 / h\right)^2}\right)^{-1} ; \quad C=\text { const. }
$$
For we then have
$$
\alpha=\frac{C}{h} \quad \text { and } \quad \beta=-\frac{C}{h} \frac{y_2}{h}\left(\sqrt{1-\left(y_2 / h\right)^2}\right)^{-1}
$$
a constant and a function of $y_2$ alone, respectively.
The first and second of Eqs. (7.78) integrate to give
$$
v_1=\frac{C}{h} y_1+f\left(y_2\right) \text { and } v_2=-\frac{C}{h} y_2+g\left(y_1\right),
$$
where $f$ and $g$ are functions to be determined. These yield $v_{1,2}=f^{\prime}\left(y_2\right)$ and $v_{2,1}=g^{\prime}\left(y_1\right)$, which combine with the third equality in (7.78) and the second in (7.81) to give $g^{\prime \prime}\left(y_1\right)=0$ and
$$
v_2=-\frac{C}{h} y_2+D y_1+E,
$$
where $D$ and $E$ are constants. Then, $v_{1,2}=2 \beta-D$, yielding
$$
v_1=C\left[y_1 / h+2 \sqrt{1-\left(y_2 / h\right)^2}\right]-D y_2+F
$$
where $F$ is another constant.
Clearly $E$ and $F$ are rigid-body translational velocities. Further, the terms involving $D$ contribute a fixed tensor to the vorticity $\mathbf{W}$, and therefore represent a rigid-body rotation. Suppressing the rigid-body terms, we thus have the velocity field
$$
v_1=C\left[y_1 / h+2 \sqrt{1-\left(y_2 / h\right)^2}\right], \quad v_2=-C y_2 / h,
$$
valid in the region defined by $\left|y_2\right| \leq h$.
Evidently $v_2=\mp C$ at $y_2= \pm h$, and so this solution simulates the squeezing, assuming $C>0$, of a layer of thickness $2 h$ between rigid walls approaching each other at the relative speed $2 C$. The solution predicts that $T_{12}=\mp k$ at $y_2= \pm h$ but does not satisfy the noslip conditions $v_1=0$ there. We also note that $\left|D_{12}\right|(=|\beta|) \rightarrow \infty$ as $\left|y_2\right| \rightarrow h$. Thus, if the material is viscoplastic, then the viscosity cannot be neglected when $y_2$ is close to $\pm h$. In this case the present solution is valid outside a thin boundary layer adjoining the walls in which Bingham's model may be used to meet a no-slip condition. See Prager's book for a discussion of the relevant boundary-layer theory. Further discussion of the present solution is given in the books by Nadai and Hill, the latter including corrections to accommodate finite-width strips.

Corresponding slip lines may be determined, if desired, by using $(7.70)_1$ to obtain
$$
d y_2 / d y_1=2 h(\sin 2 \theta) d \theta / d y_1
$$
Then, on the trajectory with unit tangent $\mathbf{t}$, where $d y_2 / d y_1=\tan \theta$, and on that with unit tangent s, where $d y_2 / d y_1=-\cot \theta$, we have
$$
y_1=h(\sin 2 \theta+2 \theta)+\text { const. } \quad \text { and } \quad y_1=h(\sin 2 \theta-2 \theta)+\text { const. }
$$
respectively. These, together with $(7.70)_1$, are the parametric equations of a pair of orthogonal cycloids.

Check back soon!

Problem 10

Consider a crystalline solid having cubic symmetry relative to $\kappa_i(p)$. Let the edges of the cube be aligned with an orthonormal basis $\left\{\mathbf{1}_i\right\}$. It is known-see, for example, the book by Green and Adkins-that the most general homogeneous quadratic strain-energy function for such a solid is a linear combination of
$$
\left(E_{11}+E_{22}+E_{33}\right)^2, \quad E_{11} E_{22}+E_{11} E_{33}+E_{22} E_{33} \text { and } E_{12}^2+E_{13}^2+E_{23}^2,
$$
where $E_{i j}=\hat{\mathbf{E}} \cdot \operatorname{Sym}\left(\mathbf{1}_i \otimes \mathbf{1}_j\right)$ are the Cartesian components of the strain $\hat{\mathbf{E}}$.
(a) Show that the general quadratic strain-energy function for a uniform cubic solid is thus expressible in the form
$$
U(\hat{\mathbf{E}})=\frac{1}{2}\left[C_1\left(E_{11}+E_{22}+E_{33}\right)^2+C_2\left(\bar{E}_{11}^2+\bar{E}_{22}^2+\bar{E}_{33}^2\right)+C_3\left(E_{12}^2+E_{13}^2+E_{23}^2\right)\right],
$$
(b) Show that $U(\hat{\mathbf{E}})$ is positive definite if and only if all $C_i>0$.
(c) Derive expressions for the Cartesian components $S_{i j}$ of the Piola-Kirchhoff stress $\hat{\mathbf{S}}=U_{\hat{\mathbf{E}}}$ relative to the basis $\left\{\mathbf{1}_i \otimes \mathbf{1}_j\right\}$.

Check back soon!
03:23

Problem 11

Use (7.52) to verify that the zero-traction condition, $\mathbf{T i}_1=\mathbf{0}$, is satisfied.
To illustrate a simple state, consider the problem of a straight, semi-infinite tractionfree crack and the tentative slip-line field sketched in Figure 7.1. This is partitioned into the three regions labeled $A, B$, and $C$, in which $A$ and $C$ are constant states and $B$ is a simple state. In region $A, T_{22}$ and $T_{12}$ vanish, and, assuming $T_{11}$ to be positive, $T_{11}=2 k$. We also have $\theta=3 \pi / 4$ and $p=-k$.

If we trace a $\beta$-line-a curve on which $\alpha$ is constant-from region $A$ into region $C$, where $\theta=\pi / 4$, and make use of $(7.89)_2$, we obtain $-p / 2 k+\pi / 4=k / 2 k+3 \pi / 4$. Thus, $p=-(1+\pi) k$ in region $C$.

Manish Jain
Manish Jain
Numerade Educator
03:23

Problem 11

Use (7.52) to verify that the zero-traction condition, $\mathrm{Ti}_1=0$, is satisfied.
To illustrate a simple state, consider the problem of a straight, semi-infinite tractionfree crack and the tentative slip-line field sketched in Figure 7.1. This is partitioned into the three regions labeled $A, B$, and $C$, in which $A$ and $C$ are constant states and $B$ is a simple state. In region $A, T_{22}$ and $T_{12}$ vanish, and, assuming $T_{11}$ to be positive, $T_{11}=2 k$. We also have $\theta=3 \pi / 4$ and $p=-k$.

If we trace a $\beta$-line-a curve on which $\alpha$ is constant-from region $A$ into region $C$, where $\theta=\pi / 4$, and make use of $(7.89)_2$, we obtain $-p / 2 k+\pi / 4=k / 2 k+3 \pi / 4$. Thus, $p=-(1+\pi) k$ in region $C$.

Manish Jain
Manish Jain
Numerade Educator

Problem 12

Use (7.52) to derive the Cartesian stress components $T_{11}=\pi k, T_{22}=$ $(\pi+2) k$, and $T_{12}=0$ in region $C$.

The state in region $B$ is similarly obtained by tracing a $\beta$-line from region $A$. Thus, $-p / 2 k+\varphi=k / 2 k+3 \pi / 4$, where $\varphi$ is the azimuthal angle in a system of plane polar
coordinates with origin at the crack tip. Thus,
$$
\mathbf{t}=\mathbf{e}_r(\varphi) \text { and } \mathbf{s}=\mathbf{e}_{\varphi}(\varphi)
$$
where $\mathbf{e}_r$ and $\mathbf{e}_{\varphi}$ are defined in Problem 7.5. Thus, in region $B$ we have
$$
\theta=\varphi, \quad p=-k(1+3 \pi / 2-2 \varphi) ; \quad \pi / 4 \leq \varphi \leq 3 \pi / 4
$$
Evidently the stress field in the vicinity of the crack tip is bounded. Note that, as no length scale is specified in this problem, the extent of the plastic regions cannot be determined.

Check back soon!

Problem 12

Use (7.52) to derive the Cartesian stress components $T_{11}=\pi k, T_{22}=$ $(\pi+2) k$, and $T_{12}=0$ in region $C$.
The state in region $B$ is similarly obtained by tracing a $\beta$-line from region $A$. Thus, $-p / 2 k+\varphi=k / 2 k+3 \pi / 4$, where $\varphi$ is the azimuthal angle in a system of plane polar
coordinates with origin at the crack tip. Thus,
$$
\mathbf{t}=\mathbf{e}_r(\varphi) \text { and } \mathbf{s}=\mathbf{e}_{\varphi}(\varphi)
$$
where $\mathbf{e}_r$ and $\mathbf{e}_{\varphi}$ are defined in Problem 7.5. Thus, in region $B$ we have
$$
\theta=\varphi, \quad p=-k(1+3 \pi / 2-2 \varphi) ; \quad \pi / 4 \leq \varphi \leq 3 \pi / 4
$$
Evidently the stress field in the vicinity of the crack tip is bounded. Note that, as no length scale is specified in this problem, the extent of the plastic regions cannot be determined.

Check back soon!
03:20

Problem 13

Use (7.52) to write the stress in the polar form $\mathbf{T}=T_{r r} \mathbf{e}_r \otimes \mathbf{e}_r+T_{\varphi \varphi} \mathbf{e}_{\varphi} \otimes$ $\mathbf{e}_{\varphi}+T_{r \varphi}\left(\mathbf{e}_r \otimes \mathbf{e}_{\varphi}+\mathbf{e}_{\varphi} \otimes \mathbf{e}_r\right)+T_{z z} \mathbf{k} \otimes \mathbf{k}$. Show that $T_{\pi r}=T_{\varphi \varphi}=T_{z z}=-p$ and $T_r \varphi=k$ in region $B$.

Himanshu Kushwaha
Himanshu Kushwaha
Numerade Educator