Menu
Home
People
Places
Arts
History
Plants & Animals
Science
Life & Culture
Technology
Reference.org
Tensor derivative (continuum mechanics)
open-in-new
<h2 id="derivatives-with-respect-to-vectors-and-second-order-tensors">Derivatives with respect to vectors and second-order tensors</h2> <p>The definitions of directional derivatives for various situations are given below. It is assumed that the functions are sufficiently smooth that derivatives can be taken. </p> <h3>Derivatives of scalar valued functions of vectors</h3> <p>Let <i>f</i>(v) be a real valued function of the vector v. Then the derivative of <i>f</i>(v) with respect to v (or at v) is the vector defined through its <a href="/facts/Dot_product/tNz8MLjT">dot product</a> with any vector u being </p><p> ∂ f ∂ v ⋅ u = D f ( v ) [ u ] = [ d d α f ( v + α u ) ] α = 0 {\displaystyle {\frac {\partial f}{\partial \mathbf {v} }}\cdot \mathbf {u} =Df(\mathbf {v} )[\mathbf {u} ]=\left[{\frac {d}{d\alpha }}~f(\mathbf {v} +\alpha ~\mathbf {u} )\right]_{\alpha =0}} </p><p>for all vectors u. The above dot product yields a scalar, and if u is a unit vector gives the directional derivative of <i>f</i> at v, in the u direction. </p><p>Properties: </p> <ol><li>If f ( v ) = f 1 ( v ) + f 2 ( v ) {\displaystyle f(\mathbf {v} )=f_{1}(\mathbf {v} )+f_{2}(\mathbf {v} )} then ∂ f ∂ v ⋅ u = ( ∂ f 1 ∂ v + ∂ f 2 ∂ v ) ⋅ u {\displaystyle {\frac {\partial f}{\partial \mathbf {v} }}\cdot \mathbf {u} =\left({\frac {\partial f_{1}}{\partial \mathbf {v} }}+{\frac {\partial f_{2}}{\partial \mathbf {v} }}\right)\cdot \mathbf {u} } </li> <li>If f ( v ) = f 1 ( v ) f 2 ( v ) {\displaystyle f(\mathbf {v} )=f_{1}(\mathbf {v} )~f_{2}(\mathbf {v} )} then ∂ f ∂ v ⋅ u = ( ∂ f 1 ∂ v ⋅ u ) f 2 ( v ) + f 1 ( v ) ( ∂ f 2 ∂ v ⋅ u ) {\displaystyle {\frac {\partial f}{\partial \mathbf {v} }}\cdot \mathbf {u} =\left({\frac {\partial f_{1}}{\partial \mathbf {v} }}\cdot \mathbf {u} \right)~f_{2}(\mathbf {v} )+f_{1}(\mathbf {v} )~\left({\frac {\partial f_{2}}{\partial \mathbf {v} }}\cdot \mathbf {u} \right)} </li> <li>If f ( v ) = f 1 ( f 2 ( v ) ) {\displaystyle f(\mathbf {v} )=f_{1}(f_{2}(\mathbf {v} ))} then ∂ f ∂ v ⋅ u = ∂ f 1 ∂ f 2 ∂ f 2 ∂ v ⋅ u {\displaystyle {\frac {\partial f}{\partial \mathbf {v} }}\cdot \mathbf {u} ={\frac {\partial f_{1}}{\partial f_{2}}}~{\frac {\partial f_{2}}{\partial \mathbf {v} }}\cdot \mathbf {u} } </li></ol> <h3>Derivatives of vector valued functions of vectors</h3> <p>Let f(v) be a vector valued function of the vector v. Then the derivative of f(v) with respect to v (or at v) is the second order tensor defined through its dot product with any vector u being </p><p> ∂ f ∂ v ⋅ u = D f ( v ) [ u ] = [ d d α f ( v + α u ) ] α = 0 {\displaystyle {\frac {\partial \mathbf {f} }{\partial \mathbf {v} }}\cdot \mathbf {u} =D\mathbf {f} (\mathbf {v} )[\mathbf {u} ]=\left[{\frac {d}{d\alpha }}~\mathbf {f} (\mathbf {v} +\alpha ~\mathbf {u} )\right]_{\alpha =0}} </p><p>for all vectors u. The above dot product yields a vector, and if u is a unit vector gives the direction derivative of f at v, in the directional u. </p><p>Properties: </p> <ol><li>If f ( v ) = f 1 ( v ) + f 2 ( v ) {\displaystyle \mathbf {f} (\mathbf {v} )=\mathbf {f} _{1}(\mathbf {v} )+\mathbf {f} _{2}(\mathbf {v} )} then ∂ f ∂ v ⋅ u = ( ∂ f 1 ∂ v + ∂ f 2 ∂ v ) ⋅ u {\displaystyle {\frac {\partial \mathbf {f} }{\partial \mathbf {v} }}\cdot \mathbf {u} =\left({\frac {\partial \mathbf {f} _{1}}{\partial \mathbf {v} }}+{\frac {\partial \mathbf {f} _{2}}{\partial \mathbf {v} }}\right)\cdot \mathbf {u} } </li> <li>If f ( v ) = f 1 ( v ) × f 2 ( v ) {\displaystyle \mathbf {f} (\mathbf {v} )=\mathbf {f} _{1}(\mathbf {v} )\times \mathbf {f} _{2}(\mathbf {v} )} then ∂ f ∂ v ⋅ u = ( ∂ f 1 ∂ v ⋅ u ) × f 2 ( v ) + f 1 ( v ) × ( ∂ f 2 ∂ v ⋅ u ) {\displaystyle {\frac {\partial \mathbf {f} }{\partial \mathbf {v} }}\cdot \mathbf {u} =\left({\frac {\partial \mathbf {f} _{1}}{\partial \mathbf {v} }}\cdot \mathbf {u} \right)\times \mathbf {f} _{2}(\mathbf {v} )+\mathbf {f} _{1}(\mathbf {v} )\times \left({\frac {\partial \mathbf {f} _{2}}{\partial \mathbf {v} }}\cdot \mathbf {u} \right)} </li> <li>If f ( v ) = f 1 ( f 2 ( v ) ) {\displaystyle \mathbf {f} (\mathbf {v} )=\mathbf {f} _{1}(\mathbf {f} _{2}(\mathbf {v} ))} then ∂ f ∂ v ⋅ u = ∂ f 1 ∂ f 2 ⋅ ( ∂ f 2 ∂ v ⋅ u ) {\displaystyle {\frac {\partial \mathbf {f} }{\partial \mathbf {v} }}\cdot \mathbf {u} ={\frac {\partial \mathbf {f} _{1}}{\partial \mathbf {f} _{2}}}\cdot \left({\frac {\partial \mathbf {f} _{2}}{\partial \mathbf {v} }}\cdot \mathbf {u} \right)} </li></ol> <h3>Derivatives of scalar valued functions of second-order tensors</h3> <p>Let f ( S ) {\displaystyle f({\boldsymbol {S}})} be a real valued function of the second order tensor S {\displaystyle {\boldsymbol {S}}} . Then the derivative of f ( S ) {\displaystyle f({\boldsymbol {S}})} with respect to S {\displaystyle {\boldsymbol {S}}} (or at S {\displaystyle {\boldsymbol {S}}} ) in the direction T {\displaystyle {\boldsymbol {T}}} is the second order tensor defined as ∂ f ∂ S : T = D f ( S ) [ T ] = [ d d α f ( S + α T ) ] α = 0 {\displaystyle {\frac {\partial f}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}=Df({\boldsymbol {S}})[{\boldsymbol {T}}]=\left[{\frac {d}{d\alpha }}~f({\boldsymbol {S}}+\alpha ~{\boldsymbol {T}})\right]_{\alpha =0}} for all second order tensors T {\displaystyle {\boldsymbol {T}}} . </p><p>Properties: </p> <ol><li>If f ( S ) = f 1 ( S ) + f 2 ( S ) {\displaystyle f({\boldsymbol {S}})=f_{1}({\boldsymbol {S}})+f_{2}({\boldsymbol {S}})} then ∂ f ∂ S : T = ( ∂ f 1 ∂ S + ∂ f 2 ∂ S ) : T {\displaystyle {\frac {\partial f}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}=\left({\frac {\partial f_{1}}{\partial {\boldsymbol {S}}}}+{\frac {\partial f_{2}}{\partial {\boldsymbol {S}}}}\right):{\boldsymbol {T}}} </li> <li>If f ( S ) = f 1 ( S ) f 2 ( S ) {\displaystyle f({\boldsymbol {S}})=f_{1}({\boldsymbol {S}})~f_{2}({\boldsymbol {S}})} then ∂ f ∂ S : T = ( ∂ f 1 ∂ S : T ) f 2 ( S ) + f 1 ( S ) ( ∂ f 2 ∂ S : T ) {\displaystyle {\frac {\partial f}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}=\left({\frac {\partial f_{1}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)~f_{2}({\boldsymbol {S}})+f_{1}({\boldsymbol {S}})~\left({\frac {\partial f_{2}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)} </li> <li>If f ( S ) = f 1 ( f 2 ( S ) ) {\displaystyle f({\boldsymbol {S}})=f_{1}(f_{2}({\boldsymbol {S}}))} then ∂ f ∂ S : T = ∂ f 1 ∂ f 2 ( ∂ f 2 ∂ S : T ) {\displaystyle {\frac {\partial f}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}={\frac {\partial f_{1}}{\partial f_{2}}}~\left({\frac {\partial f_{2}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)} </li></ol> <h3>Derivatives of tensor valued functions of second-order tensors</h3> <p>Let F ( S ) {\displaystyle {\boldsymbol {F}}({\boldsymbol {S}})} be a second order tensor valued function of the second order tensor S {\displaystyle {\boldsymbol {S}}} . Then the derivative of F ( S ) {\displaystyle {\boldsymbol {F}}({\boldsymbol {S}})} with respect to S {\displaystyle {\boldsymbol {S}}} (or at S {\displaystyle {\boldsymbol {S}}} ) in the direction T {\displaystyle {\boldsymbol {T}}} is the fourth order tensor defined as ∂ F ∂ S : T = D F ( S ) [ T ] = [ d d α F ( S + α T ) ] α = 0 {\displaystyle {\frac {\partial {\boldsymbol {F}}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}=D{\boldsymbol {F}}({\boldsymbol {S}})[{\boldsymbol {T}}]=\left[{\frac {d}{d\alpha }}~{\boldsymbol {F}}({\boldsymbol {S}}+\alpha ~{\boldsymbol {T}})\right]_{\alpha =0}} for all second order tensors T {\displaystyle {\boldsymbol {T}}} . </p><p>Properties: </p> <ol><li>If F ( S ) = F 1 ( S ) + F 2 ( S ) {\displaystyle {\boldsymbol {F}}({\boldsymbol {S}})={\boldsymbol {F}}_{1}({\boldsymbol {S}})+{\boldsymbol {F}}_{2}({\boldsymbol {S}})} then ∂ F ∂ S : T = ( ∂ F 1 ∂ S + ∂ F 2 ∂ S ) : T {\displaystyle {\frac {\partial {\boldsymbol {F}}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}=\left({\frac {\partial {\boldsymbol {F}}_{1}}{\partial {\boldsymbol {S}}}}+{\frac {\partial {\boldsymbol {F}}_{2}}{\partial {\boldsymbol {S}}}}\right):{\boldsymbol {T}}} </li> <li>If F ( S ) = F 1 ( S ) ⋅ F 2 ( S ) {\displaystyle {\boldsymbol {F}}({\boldsymbol {S}})={\boldsymbol {F}}_{1}({\boldsymbol {S}})\cdot {\boldsymbol {F}}_{2}({\boldsymbol {S}})} then ∂ F ∂ S : T = ( ∂ F 1 ∂ S : T ) ⋅ F 2 ( S ) + F 1 ( S ) ⋅ ( ∂ F 2 ∂ S : T ) {\displaystyle {\frac {\partial {\boldsymbol {F}}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}=\left({\frac {\partial {\boldsymbol {F}}_{1}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)\cdot {\boldsymbol {F}}_{2}({\boldsymbol {S}})+{\boldsymbol {F}}_{1}({\boldsymbol {S}})\cdot \left({\frac {\partial {\boldsymbol {F}}_{2}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)} </li> <li>If F ( S ) = F 1 ( F 2 ( S ) ) {\displaystyle {\boldsymbol {F}}({\boldsymbol {S}})={\boldsymbol {F}}_{1}({\boldsymbol {F}}_{2}({\boldsymbol {S}}))} then ∂ F ∂ S : T = ∂ F 1 ∂ F 2 : ( ∂ F 2 ∂ S : T ) {\displaystyle {\frac {\partial {\boldsymbol {F}}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}={\frac {\partial {\boldsymbol {F}}_{1}}{\partial {\boldsymbol {F}}_{2}}}:\left({\frac {\partial {\boldsymbol {F}}_{2}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)} </li> <li>If f ( S ) = f 1 ( F 2 ( S ) ) {\displaystyle f({\boldsymbol {S}})=f_{1}({\boldsymbol {F}}_{2}({\boldsymbol {S}}))} then ∂ f ∂ S : T = ∂ f 1 ∂ F 2 : ( ∂ F 2 ∂ S : T ) {\displaystyle {\frac {\partial f}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}={\frac {\partial f_{1}}{\partial {\boldsymbol {F}}_{2}}}:\left({\frac {\partial {\boldsymbol {F}}_{2}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)} </li></ol> <h2 id="gradient-of-a-tensor-field">Gradient of a tensor field</h2> <p>The <a href="/facts/Gradient/TsR7uukj">gradient</a>, ∇ T {\displaystyle {\boldsymbol {\nabla }}{\boldsymbol {T}}} , of a tensor field T ( x ) {\displaystyle {\boldsymbol {T}}(\mathbf {x} )} in the direction of an arbitrary constant vector c is defined as: ∇ T ⋅ c = lim α → 0 d d α T ( x + α c ) {\displaystyle {\boldsymbol {\nabla }}{\boldsymbol {T}}\cdot \mathbf {c} =\lim _{\alpha \rightarrow 0}\quad {\cfrac {d}{d\alpha }}~{\boldsymbol {T}}(\mathbf {x} +\alpha \mathbf {c} )} The gradient of a tensor field of order <i>n</i> is a tensor field of order <i>n</i>+1. </p> <h3>Cartesian coordinates</h3> <p class="note">Note: the <a href="/facts/Einstein_summation_convention/UXwS43wU">Einstein summation convention</a> of summing on repeated indices is used below.</p> <p>If e 1 , e 2 , e 3 {\displaystyle \mathbf {e} _{1},\mathbf {e} _{2},\mathbf {e} _{3}} are the basis vectors in a <a href="/facts/Cartesian_coordinate/lkTqvMmB">Cartesian coordinate</a> system, with coordinates of points denoted by ( x 1 , x 2 , x 3 {\displaystyle x_{1},x_{2},x_{3}} ), then the gradient of the tensor field T {\displaystyle {\boldsymbol {T}}} is given by ∇ T = ∂ T ∂ x i ⊗ e i {\displaystyle {\boldsymbol {\nabla }}{\boldsymbol {T}}={\cfrac {\partial {\boldsymbol {T}}}{\partial x_{i}}}\otimes \mathbf {e} _{i}} </p> Proof <p>The vectors x and c can be written as x = x i e i {\displaystyle \mathbf {x} =x_{i}~\mathbf {e} _{i}} and c = c i e i {\displaystyle \mathbf {c} =c_{i}~\mathbf {e} _{i}} . Let y := x + αc. In that case the gradient is given by ∇ T ⋅ c = d d α T ( x 1 + α c 1 , x 2 + α c 2 , x 3 + α c 3 ) | α = 0 ≡ d d α T ( y 1 , y 2 , y 3 ) | α = 0 = [ ∂ T ∂ y 1 ∂ y 1 ∂ α + ∂ T ∂ y 2 ∂ y 2 ∂ α + ∂ T ∂ y 3 ∂ y 3 ∂ α ] α = 0 = [ ∂ T ∂ y 1 c 1 + ∂ T ∂ y 2 c 2 + ∂ T ∂ y 3 c 3 ] α = 0 = ∂ T ∂ x 1 c 1 + ∂ T ∂ x 2 c 2 + ∂ T ∂ x 3 c 3 ≡ ∂ T ∂ x i c i = ∂ T ∂ x i ( e i ⋅ c ) = [ ∂ T ∂ x i ⊗ e i ] ⋅ c ◻ {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}{\boldsymbol {T}}\cdot \mathbf {c} &=\left.{\cfrac {d}{d\alpha }}~{\boldsymbol {T}}(x_{1}+\alpha c_{1},x_{2}+\alpha c_{2},x_{3}+\alpha c_{3})\right|_{\alpha =0}\equiv \left.{\cfrac {d}{d\alpha }}~{\boldsymbol {T}}(y_{1},y_{2},y_{3})\right|_{\alpha =0}\\&=\left[{\cfrac {\partial {\boldsymbol {T}}}{\partial y_{1}}}~{\cfrac {\partial y_{1}}{\partial \alpha }}+{\cfrac {\partial {\boldsymbol {T}}}{\partial y_{2}}}~{\cfrac {\partial y_{2}}{\partial \alpha }}+{\cfrac {\partial {\boldsymbol {T}}}{\partial y_{3}}}~{\cfrac {\partial y_{3}}{\partial \alpha }}\right]_{\alpha =0}=\left[{\cfrac {\partial {\boldsymbol {T}}}{\partial y_{1}}}~c_{1}+{\cfrac {\partial {\boldsymbol {T}}}{\partial y_{2}}}~c_{2}+{\cfrac {\partial {\boldsymbol {T}}}{\partial y_{3}}}~c_{3}\right]_{\alpha =0}\\&={\cfrac {\partial {\boldsymbol {T}}}{\partial x_{1}}}~c_{1}+{\cfrac {\partial {\boldsymbol {T}}}{\partial x_{2}}}~c_{2}+{\cfrac {\partial {\boldsymbol {T}}}{\partial x_{3}}}~c_{3}\equiv {\cfrac {\partial {\boldsymbol {T}}}{\partial x_{i}}}~c_{i}={\cfrac {\partial {\boldsymbol {T}}}{\partial x_{i}}}~(\mathbf {e} _{i}\cdot \mathbf {c} )=\left[{\cfrac {\partial {\boldsymbol {T}}}{\partial x_{i}}}\otimes \mathbf {e} _{i}\right]\cdot \mathbf {c} \qquad \square \end{aligned}}} </p> <p>Since the basis vectors do not vary in a Cartesian coordinate system we have the following relations for the gradients of a scalar field ϕ {\displaystyle \phi } , a vector field v, and a second-order tensor field S {\displaystyle {\boldsymbol {S}}} . ∇ ϕ = ∂ ϕ ∂ x i e i = ϕ , i e i ∇ v = ∂ ( v j e j ) ∂ x i ⊗ e i = ∂ v j ∂ x i e j ⊗ e i = v j , i e j ⊗ e i ∇ S = ∂ ( S j k e j ⊗ e k ) ∂ x i ⊗ e i = ∂ S j k ∂ x i e j ⊗ e k ⊗ e i = S j k , i e j ⊗ e k ⊗ e i {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\phi &={\cfrac {\partial \phi }{\partial x_{i}}}~\mathbf {e} _{i}=\phi _{,i}~\mathbf {e} _{i}\\{\boldsymbol {\nabla }}\mathbf {v} &={\cfrac {\partial (v_{j}\mathbf {e} _{j})}{\partial x_{i}}}\otimes \mathbf {e} _{i}={\cfrac {\partial v_{j}}{\partial x_{i}}}~\mathbf {e} _{j}\otimes \mathbf {e} _{i}=v_{j,i}~\mathbf {e} _{j}\otimes \mathbf {e} _{i}\\{\boldsymbol {\nabla }}{\boldsymbol {S}}&={\cfrac {\partial (S_{jk}\mathbf {e} _{j}\otimes \mathbf {e} _{k})}{\partial x_{i}}}\otimes \mathbf {e} _{i}={\cfrac {\partial S_{jk}}{\partial x_{i}}}~\mathbf {e} _{j}\otimes \mathbf {e} _{k}\otimes \mathbf {e} _{i}=S_{jk,i}~\mathbf {e} _{j}\otimes \mathbf {e} _{k}\otimes \mathbf {e} _{i}\end{aligned}}} </p> <h3>Curvilinear coordinates</h3> <p class="note">Main article: <a href="/facts/Tensors_in_curvilinear_coordinates/CC55RJpm">Tensors in curvilinear coordinates</a></p> <p class="note">Note: the <a href="/facts/Einstein_summation_convention/UXwS43wU">Einstein summation convention</a> of summing on repeated indices is used below.</p> <p>If g 1 , g 2 , g 3 {\displaystyle \mathbf {g} ^{1},\mathbf {g} ^{2},\mathbf {g} ^{3}} are the <a href="/facts/Covariance_and_contravariance_of_vectors/s3HGMkhz">contravariant</a> <a href="/facts/Basis_vector/89IPoN6c">basis vectors</a> in a <a href="/facts/Curvilinear_coordinates/RQD7Cajd">curvilinear coordinate</a> system, with coordinates of points denoted by ( ξ 1 , ξ 2 , ξ 3 {\displaystyle \xi ^{1},\xi ^{2},\xi ^{3}} ), then the gradient of the tensor field T {\displaystyle {\boldsymbol {T}}} is given by<a class="footnote-ref" id="fnref:3" href="#fn:3"><sup>3</sup></a> ∇ T = ∂ T ∂ ξ i ⊗ g i {\displaystyle {\boldsymbol {\nabla }}{\boldsymbol {T}}={\frac {\partial {\boldsymbol {T}}}{\partial \xi ^{i}}}\otimes \mathbf {g} ^{i}} </p><p>From this definition we have the following relations for the gradients of a scalar field ϕ {\displaystyle \phi } , a vector field v, and a second-order tensor field S {\displaystyle {\boldsymbol {S}}} . ∇ ϕ = ∂ ϕ ∂ ξ i g i ∇ v = ∂ ( v j g j ) ∂ ξ i ⊗ g i = ( ∂ v j ∂ ξ i + v k Γ i k j ) g j ⊗ g i = ( ∂ v j ∂ ξ i − v k Γ i j k ) g j ⊗ g i ∇ S = ∂ ( S j k g j ⊗ g k ) ∂ ξ i ⊗ g i = ( ∂ S j k ∂ ξ i − S l k Γ i j l − S j l Γ i k l ) g j ⊗ g k ⊗ g i {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\phi &={\frac {\partial \phi }{\partial \xi ^{i}}}~\mathbf {g} ^{i}\\[1.2ex]{\boldsymbol {\nabla }}\mathbf {v} &={\frac {\partial \left(v^{j}\mathbf {g} _{j}\right)}{\partial \xi ^{i}}}\otimes \mathbf {g} ^{i}\\&=\left({\frac {\partial v^{j}}{\partial \xi ^{i}}}+v^{k}~\Gamma _{ik}^{j}\right)~\mathbf {g} _{j}\otimes \mathbf {g} ^{i}=\left({\frac {\partial v_{j}}{\partial \xi ^{i}}}-v_{k}~\Gamma _{ij}^{k}\right)~\mathbf {g} ^{j}\otimes \mathbf {g} ^{i}\\[1.2ex]{\boldsymbol {\nabla }}{\boldsymbol {S}}&={\frac {\partial \left(S_{jk}~\mathbf {g} ^{j}\otimes \mathbf {g} ^{k}\right)}{\partial \xi ^{i}}}\otimes \mathbf {g} ^{i}\\&=\left({\frac {\partial S_{jk}}{\partial \xi _{i}}}-S_{lk}~\Gamma _{ij}^{l}-S_{jl}~\Gamma _{ik}^{l}\right)~\mathbf {g} ^{j}\otimes \mathbf {g} ^{k}\otimes \mathbf {g} ^{i}\end{aligned}}} </p><p>where the <a href="/facts/Christoffel_symbol/uCyd7RWX">Christoffel symbol</a> Γ i j k {\displaystyle \Gamma _{ij}^{k}} is defined using Γ i j k g k = ∂ g i ∂ ξ j ⟹ Γ i j k = ∂ g i ∂ ξ j ⋅ g k = − g i ⋅ ∂ g k ∂ ξ j {\displaystyle \Gamma _{ij}^{k}~\mathbf {g} _{k}={\frac {\partial \mathbf {g} _{i}}{\partial \xi ^{j}}}\quad \implies \quad \Gamma _{ij}^{k}={\frac {\partial \mathbf {g} _{i}}{\partial \xi ^{j}}}\cdot \mathbf {g} ^{k}=-\mathbf {g} _{i}\cdot {\frac {\partial \mathbf {g} ^{k}}{\partial \xi ^{j}}}} </p> <h4>Cylindrical polar coordinates</h4> <p>In <a href="/facts/Cylindrical_coordinate/1Vc6lhxD">cylindrical coordinates</a>, the gradient is given by ∇ ϕ = ∂ ϕ ∂ r e r + 1 r ∂ ϕ ∂ θ e θ + ∂ ϕ ∂ z e z {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\phi ={}\quad &{\frac {\partial \phi }{\partial r}}~\mathbf {e} _{r}+{\frac {1}{r}}~{\frac {\partial \phi }{\partial \theta }}~\mathbf {e} _{\theta }+{\frac {\partial \phi }{\partial z}}~\mathbf {e} _{z}\\\end{aligned}}} </p><p> ∇ v = ∂ v r ∂ r e r ⊗ e r + 1 r ( ∂ v r ∂ θ − v θ ) e r ⊗ e θ + ∂ v r ∂ z e r ⊗ e z + ∂ v θ ∂ r e θ ⊗ e r + 1 r ( ∂ v θ ∂ θ + v r ) e θ ⊗ e θ + ∂ v θ ∂ z e θ ⊗ e z + ∂ v z ∂ r e z ⊗ e r + 1 r ∂ v z ∂ θ e z ⊗ e θ + ∂ v z ∂ z e z ⊗ e z {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\mathbf {v} ={}\quad &{\frac {\partial v_{r}}{\partial r}}~\mathbf {e} _{r}\otimes \mathbf {e} _{r}+{\frac {1}{r}}\left({\frac {\partial v_{r}}{\partial \theta }}-v_{\theta }\right)~\mathbf {e} _{r}\otimes \mathbf {e} _{\theta }+{\frac {\partial v_{r}}{\partial z}}~\mathbf {e} _{r}\otimes \mathbf {e} _{z}\\{}+{}&{\frac {\partial v_{\theta }}{\partial r}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{r}+{\frac {1}{r}}\left({\frac {\partial v_{\theta }}{\partial \theta }}+v_{r}\right)~\mathbf {e} _{\theta }\otimes \mathbf {e} _{\theta }+{\frac {\partial v_{\theta }}{\partial z}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{z}\\{}+{}&{\frac {\partial v_{z}}{\partial r}}~\mathbf {e} _{z}\otimes \mathbf {e} _{r}+{\frac {1}{r}}{\frac {\partial v_{z}}{\partial \theta }}~\mathbf {e} _{z}\otimes \mathbf {e} _{\theta }+{\frac {\partial v_{z}}{\partial z}}~\mathbf {e} _{z}\otimes \mathbf {e} _{z}\\\end{aligned}}} </p><p> ∇ S = ∂ S r r ∂ r e r ⊗ e r ⊗ e r + ∂ S r r ∂ z e r ⊗ e r ⊗ e z + 1 r [ ∂ S r r ∂ θ − ( S θ r + S r θ ) ] e r ⊗ e r ⊗ e θ + ∂ S r θ ∂ r e r ⊗ e θ ⊗ e r + ∂ S r θ ∂ z e r ⊗ e θ ⊗ e z + 1 r [ ∂ S r θ ∂ θ + ( S r r − S θ θ ) ] e r ⊗ e θ ⊗ e θ + ∂ S r z ∂ r e r ⊗ e z ⊗ e r + ∂ S r z ∂ z e r ⊗ e z ⊗ e z + 1 r [ ∂ S r z ∂ θ − S θ z ] e r ⊗ e z ⊗ e θ + ∂ S θ r ∂ r e θ ⊗ e r ⊗ e r + ∂ S θ r ∂ z e θ ⊗ e r ⊗ e z + 1 r [ ∂ S θ r ∂ θ + ( S r r − S θ θ ) ] e θ ⊗ e r ⊗ e θ + ∂ S θ θ ∂ r e θ ⊗ e θ ⊗ e r + ∂ S θ θ ∂ z e θ ⊗ e θ ⊗ e z + 1 r [ ∂ S θ θ ∂ θ + ( S r θ + S θ r ) ] e θ ⊗ e θ ⊗ e θ + ∂ S θ z ∂ r e θ ⊗ e z ⊗ e r + ∂ S θ z ∂ z e θ ⊗ e z ⊗ e z + 1 r [ ∂ S θ z ∂ θ + S r z ] e θ ⊗ e z ⊗ e θ + ∂ S z r ∂ r e z ⊗ e r ⊗ e r + ∂ S z r ∂ z e z ⊗ e r ⊗ e z + 1 r [ ∂ S z r ∂ θ − S z θ ] e z ⊗ e r ⊗ e θ + ∂ S z θ ∂ r e z ⊗ e θ ⊗ e r + ∂ S z θ ∂ z e z ⊗ e θ ⊗ e z + 1 r [ ∂ S z θ ∂ θ + S z r ] e z ⊗ e θ ⊗ e θ + ∂ S z z ∂ r e z ⊗ e z ⊗ e r + ∂ S z z ∂ z e z ⊗ e z ⊗ e z + 1 r ∂ S z z ∂ θ e z ⊗ e z ⊗ e θ {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}{\boldsymbol {S}}={}\quad &{\frac {\partial S_{rr}}{\partial r}}~\mathbf {e} _{r}\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{r}+{\frac {\partial S_{rr}}{\partial z}}~\mathbf {e} _{r}\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{rr}}{\partial \theta }}-(S_{\theta r}+S_{r\theta })\right]~\mathbf {e} _{r}\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{r\theta }}{\partial r}}~\mathbf {e} _{r}\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{r}+{\frac {\partial S_{r\theta }}{\partial z}}~\mathbf {e} _{r}\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{r\theta }}{\partial \theta }}+(S_{rr}-S_{\theta \theta })\right]~\mathbf {e} _{r}\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{rz}}{\partial r}}~\mathbf {e} _{r}\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{r}+{\frac {\partial S_{rz}}{\partial z}}~\mathbf {e} _{r}\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{rz}}{\partial \theta }}-S_{\theta z}\right]~\mathbf {e} _{r}\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{\theta r}}{\partial r}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{r}+{\frac {\partial S_{\theta r}}{\partial z}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{\theta r}}{\partial \theta }}+(S_{rr}-S_{\theta \theta })\right]~\mathbf {e} _{\theta }\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{\theta \theta }}{\partial r}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{r}+{\frac {\partial S_{\theta \theta }}{\partial z}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{\theta \theta }}{\partial \theta }}+(S_{r\theta }+S_{\theta r})\right]~\mathbf {e} _{\theta }\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{\theta z}}{\partial r}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{r}+{\frac {\partial S_{\theta z}}{\partial z}}~\mathbf {e} _{\theta }\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{\theta z}}{\partial \theta }}+S_{rz}\right]~\mathbf {e} _{\theta }\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{zr}}{\partial r}}~\mathbf {e} _{z}\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{r}+{\frac {\partial S_{zr}}{\partial z}}~\mathbf {e} _{z}\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{zr}}{\partial \theta }}-S_{z\theta }\right]~\mathbf {e} _{z}\otimes \mathbf {e} _{r}\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{z\theta }}{\partial r}}~\mathbf {e} _{z}\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{r}+{\frac {\partial S_{z\theta }}{\partial z}}~\mathbf {e} _{z}\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{z}+{\frac {1}{r}}\left[{\frac {\partial S_{z\theta }}{\partial \theta }}+S_{zr}\right]~\mathbf {e} _{z}\otimes \mathbf {e} _{\theta }\otimes \mathbf {e} _{\theta }\\{}+{}&{\frac {\partial S_{zz}}{\partial r}}~\mathbf {e} _{z}\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{r}+{\frac {\partial S_{zz}}{\partial z}}~\mathbf {e} _{z}\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{z}+{\frac {1}{r}}~{\frac {\partial S_{zz}}{\partial \theta }}~\mathbf {e} _{z}\otimes \mathbf {e} _{z}\otimes \mathbf {e} _{\theta }\end{aligned}}} </p> <h2 id="divergence-of-a-tensor-field">Divergence of a tensor field</h2> <p>The <a href="/facts/Divergence/zpwwkFoU">divergence</a> of a tensor field T ( x ) {\displaystyle {\boldsymbol {T}}(\mathbf {x} )} is defined using the recursive relation ( ∇ ⋅ T ) ⋅ c = ∇ ⋅ ( c ⋅ T T ) ; ∇ ⋅ v = tr ( ∇ v ) {\displaystyle ({\boldsymbol {\nabla }}\cdot {\boldsymbol {T}})\cdot \mathbf {c} ={\boldsymbol {\nabla }}\cdot \left(\mathbf {c} \cdot {\boldsymbol {T}}^{\textsf {T}}\right)~;\qquad {\boldsymbol {\nabla }}\cdot \mathbf {v} ={\text{tr}}({\boldsymbol {\nabla }}\mathbf {v} )} </p><p>where c is an arbitrary constant vector and v is a vector field. If T {\displaystyle {\boldsymbol {T}}} is a tensor field of order <i>n</i> > 1 then the divergence of the field is a tensor of order <i>n</i>− 1. </p> <h3>Cartesian coordinates</h3> <p class="note">Note: the <a href="/facts/Einstein_summation_convention/UXwS43wU">Einstein summation convention</a> of summing on repeated indices is used below.</p> <p>In a Cartesian coordinate system we have the following relations for a vector field v and a second-order tensor field S {\displaystyle {\boldsymbol {S}}} . ∇ ⋅ v = ∂ v i ∂ x i = v i , i ∇ ⋅ S = ∂ S i k ∂ x i e k = S i k , i e k {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\cdot \mathbf {v} &={\frac {\partial v_{i}}{\partial x_{i}}}=v_{i,i}\\{\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}&={\frac {\partial S_{ik}}{\partial x_{i}}}~\mathbf {e} _{k}=S_{ik,i}~\mathbf {e} _{k}\end{aligned}}} </p><p>where <a href="/facts/Ricci_calculus/We4yudeq">tensor index notation</a> for partial derivatives is used in the rightmost expressions. Note that ∇ ⋅ S ≠ ∇ ⋅ S T . {\displaystyle {\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}\neq {\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}^{\textsf {T}}.} </p><p>For a symmetric second-order tensor, the divergence is also often written as<a class="footnote-ref" id="fnref:4" href="#fn:4"><sup>4</sup></a> </p><p> ∇ ⋅ S = ∂ S k i ∂ x i e k = S k i , i e k {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}&={\cfrac {\partial S_{ki}}{\partial x_{i}}}~\mathbf {e} _{k}=S_{ki,i}~\mathbf {e} _{k}\end{aligned}}} </p><p>The above expression is sometimes used as the definition of ∇ ⋅ S {\displaystyle {\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}} in Cartesian component form (often also written as div S {\displaystyle \operatorname {div} {\boldsymbol {S}}} ). Note that such a definition is not consistent with the rest of this article (see the section on curvilinear co-ordinates). </p><p>The difference stems from whether the differentiation is performed with respect to the rows or columns of S {\displaystyle {\boldsymbol {S}}} , and is conventional. This is demonstrated by an example. In a Cartesian coordinate system the second order tensor (matrix) S {\displaystyle \mathbf {S} } is the gradient of a vector function v {\displaystyle \mathbf {v} } . </p><p> ∇ ⋅ ( ∇ v ) = ∇ ⋅ ( v i , j e i ⊗ e j ) = v i , j i e i ⋅ e i ⊗ e j = ( ∇ ⋅ v ) , j e j = ∇ ( ∇ ⋅ v ) ∇ ⋅ [ ( ∇ v ) T ] = ∇ ⋅ ( v j , i e i ⊗ e j ) = v j , i i e i ⋅ e i ⊗ e j = ∇ 2 v j e j = ∇ 2 v {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\cdot \left({\boldsymbol {\nabla }}\mathbf {v} \right)&={\boldsymbol {\nabla }}\cdot \left(v_{i,j}~\mathbf {e} _{i}\otimes \mathbf {e} _{j}\right)=v_{i,ji}~\mathbf {e} _{i}\cdot \mathbf {e} _{i}\otimes \mathbf {e} _{j}=\left({\boldsymbol {\nabla }}\cdot \mathbf {v} \right)_{,j}~\mathbf {e} _{j}={\boldsymbol {\nabla }}\left({\boldsymbol {\nabla }}\cdot \mathbf {v} \right)\\{\boldsymbol {\nabla }}\cdot \left[\left({\boldsymbol {\nabla }}\mathbf {v} \right)^{\textsf {T}}\right]&={\boldsymbol {\nabla }}\cdot \left(v_{j,i}~\mathbf {e} _{i}\otimes \mathbf {e} _{j}\right)=v_{j,ii}~\mathbf {e} _{i}\cdot \mathbf {e} _{i}\otimes \mathbf {e} _{j}={\boldsymbol {\nabla }}^{2}v_{j}~\mathbf {e} _{j}={\boldsymbol {\nabla }}^{2}\mathbf {v} \end{aligned}}} </p><p>The last equation is equivalent to the alternative definition / interpretation<a class="footnote-ref" id="fnref:5" href="#fn:5"><sup>5</sup></a> </p><p> ( ∇ ⋅ ) alt ( ∇ v ) = ( ∇ ⋅ ) alt ( v i , j e i ⊗ e j ) = v i , j j e i ⊗ e j ⋅ e j = ∇ 2 v i e i = ∇ 2 v {\displaystyle {\begin{aligned}\left({\boldsymbol {\nabla }}\cdot \right)_{\text{alt}}\left({\boldsymbol {\nabla }}\mathbf {v} \right)=\left({\boldsymbol {\nabla }}\cdot \right)_{\text{alt}}\left(v_{i,j}~\mathbf {e} _{i}\otimes \mathbf {e} _{j}\right)=v_{i,jj}~\mathbf {e} _{i}\otimes \mathbf {e} _{j}\cdot \mathbf {e} _{j}={\boldsymbol {\nabla }}^{2}v_{i}~\mathbf {e} _{i}={\boldsymbol {\nabla }}^{2}\mathbf {v} \end{aligned}}} </p> <h3>Curvilinear coordinates</h3> <p class="note">Main article: <a href="/facts/Tensors_in_curvilinear_coordinates/CC55RJpm">Tensors in curvilinear coordinates</a></p> <p class="note">Note: the <a href="/facts/Einstein_summation_convention/UXwS43wU">Einstein summation convention</a> of summing on repeated indices is used below.</p> <p>In curvilinear coordinates, the divergences of a vector field v and a second-order tensor field S {\displaystyle {\boldsymbol {S}}} are ∇ ⋅ v = ( ∂ v i ∂ ξ i + v k Γ i k i ) ∇ ⋅ S = ( ∂ S i k ∂ ξ i − S l k Γ i i l − S i l Γ i k l ) g k {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\cdot \mathbf {v} &=\left({\cfrac {\partial v^{i}}{\partial \xi ^{i}}}+v^{k}~\Gamma _{ik}^{i}\right)\\{\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}&=\left({\cfrac {\partial S_{ik}}{\partial \xi _{i}}}-S_{lk}~\Gamma _{ii}^{l}-S_{il}~\Gamma _{ik}^{l}\right)~\mathbf {g} ^{k}\end{aligned}}} </p><p>More generally, ∇ ⋅ S = [ ∂ S i j ∂ q k − Γ k i l S l j − Γ k j l S i l ] g i k b j = [ ∂ S i j ∂ q i + Γ i l i S l j + Γ i l j S i l ] b j = [ ∂ S j i ∂ q i + Γ i l i S j l − Γ i j l S l i ] b j = [ ∂ S i j ∂ q k − Γ i k l S l j + Γ k l j S i l ] g i k b j {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}&=\left[{\cfrac {\partial S_{ij}}{\partial q^{k}}}-\Gamma _{ki}^{l}~S_{lj}-\Gamma _{kj}^{l}~S_{il}\right]~g^{ik}~\mathbf {b} ^{j}\\[8pt]&=\left[{\cfrac {\partial S^{ij}}{\partial q^{i}}}+\Gamma _{il}^{i}~S^{lj}+\Gamma _{il}^{j}~S^{il}\right]~\mathbf {b} _{j}\\[8pt]&=\left[{\cfrac {\partial S_{~j}^{i}}{\partial q^{i}}}+\Gamma _{il}^{i}~S_{~j}^{l}-\Gamma _{ij}^{l}~S_{~l}^{i}\right]~\mathbf {b} ^{j}\\[8pt]&=\left[{\cfrac {\partial S_{i}^{~j}}{\partial q^{k}}}-\Gamma _{ik}^{l}~S_{l}^{~j}+\Gamma _{kl}^{j}~S_{i}^{~l}\right]~g^{ik}~\mathbf {b} _{j}\end{aligned}}} </p> <h4>Cylindrical polar coordinates</h4> <p>In <a href="/facts/Cylindrical_coordinates/1Vc6lhxD">cylindrical polar coordinates</a> ∇ ⋅ v = ∂ v r ∂ r + 1 r ( ∂ v θ ∂ θ + v r ) + ∂ v z ∂ z ∇ ⋅ S = ∂ S r r ∂ r e r + ∂ S r θ ∂ r e θ + ∂ S r z ∂ r e z + 1 r [ ∂ S θ r ∂ θ + ( S r r − S θ θ ) ] e r + 1 r [ ∂ S θ θ ∂ θ + ( S r θ + S θ r ) ] e θ + 1 r [ ∂ S θ z ∂ θ + S r z ] e z + ∂ S z r ∂ z e r + ∂ S z θ ∂ z e θ + ∂ S z z ∂ z e z {\displaystyle {\begin{aligned}{\boldsymbol {\nabla }}\cdot \mathbf {v} =\quad &{\frac {\partial v_{r}}{\partial r}}+{\frac {1}{r}}\left({\frac {\partial v_{\theta }}{\partial \theta }}+v_{r}\right)+{\frac {\partial v_{z}}{\partial z}}\\{\boldsymbol {\nabla }}\cdot {\boldsymbol {S}}=\quad &{\frac {\partial S_{rr}}{\partial r}}~\mathbf {e} _{r}+{\frac {\partial S_{r\theta }}{\partial r}}~\mathbf {e} _{\theta }+{\frac {\partial S_{rz}}{\partial r}}~\mathbf {e} _{z}\\{}+{}&{\frac {1}{r}}\left[{\frac {\partial S_{\theta r}}{\partial \theta }}+(S_{rr}-S_{\theta \theta })\right]~\mathbf {e} _{r}+{\frac {1}{r}}\left[{\frac {\partial S_{\theta \theta }}{\partial \theta }}+(S_{r\theta }+S_{\theta r})\right]~\mathbf {e} _{\theta }+{\frac {1}{r}}\left[{\frac {\partial S_{\theta z}}{\partial \theta }}+S_{rz}\right]~\mathbf {e} _{z}\\{}+{}&{\frac {\partial S_{zr}}{\partial z}}~\mathbf {e} _{r}+{\frac {\partial S_{z\theta }}{\partial z}}~\mathbf {e} _{\theta }+{\frac {\partial S_{zz}}{\partial z}}~\mathbf {e} _{z}\end{aligned}}} </p> <h2 id="curl-of-a-tensor-field">Curl of a tensor field</h2> <p>The <a href="/facts/Curl_(mathematics)/GuS8dc7o">curl</a> of an order-<i>n</i> > 1 tensor field T ( x ) {\displaystyle {\boldsymbol {T}}(\mathbf {x} )} is also defined using the recursive relation ( ∇ × T ) ⋅ c = ∇ × ( c ⋅ T ) ; ( ∇ × v ) ⋅ c = ∇ ⋅ ( v × c ) {\displaystyle ({\boldsymbol {\nabla }}\times {\boldsymbol {T}})\cdot \mathbf {c} ={\boldsymbol {\nabla }}\times (\mathbf {c} \cdot {\boldsymbol {T}})~;\qquad ({\boldsymbol {\nabla }}\times \mathbf {v} )\cdot \mathbf {c} ={\boldsymbol {\nabla }}\cdot (\mathbf {v} \times \mathbf {c} )} where c is an arbitrary constant vector and v is a vector field. </p> <h3>Curl of a first-order tensor (vector) field</h3> <p>Consider a vector field v and an arbitrary constant vector c. In index notation, the cross product is given by v × c = ε i j k v j c k e i {\displaystyle \mathbf {v} \times \mathbf {c} =\varepsilon _{ijk}~v_{j}~c_{k}~\mathbf {e} _{i}} where ε i j k {\displaystyle \varepsilon _{ijk}} is the <a href="/facts/Permutation_symbol/WBKUA3Tw">permutation symbol</a>, otherwise known as the Levi-Civita symbol. Then, ∇ ⋅ ( v × c ) = ε i j k v j , i c k = ( ε i j k v j , i e k ) ⋅ c = ( ∇ × v ) ⋅ c {\displaystyle {\boldsymbol {\nabla }}\cdot (\mathbf {v} \times \mathbf {c} )=\varepsilon _{ijk}~v_{j,i}~c_{k}=(\varepsilon _{ijk}~v_{j,i}~\mathbf {e} _{k})\cdot \mathbf {c} =({\boldsymbol {\nabla }}\times \mathbf {v} )\cdot \mathbf {c} } Therefore, ∇ × v = ε i j k v j , i e k {\displaystyle {\boldsymbol {\nabla }}\times \mathbf {v} =\varepsilon _{ijk}~v_{j,i}~\mathbf {e} _{k}} </p> <h3>Curl of a second-order tensor field</h3> <p>For a second-order tensor S {\displaystyle {\boldsymbol {S}}} c ⋅ S = c m S m j e j {\displaystyle \mathbf {c} \cdot {\boldsymbol {S}}=c_{m}~S_{mj}~\mathbf {e} _{j}} Hence, using the definition of the curl of a first-order tensor field, ∇ × ( c ⋅ S ) = ε i j k c m S m j , i e k = ( ε i j k S m j , i e k ⊗ e m ) ⋅ c = ( ∇ × S ) ⋅ c {\displaystyle {\boldsymbol {\nabla }}\times (\mathbf {c} \cdot {\boldsymbol {S}})=\varepsilon _{ijk}~c_{m}~S_{mj,i}~\mathbf {e} _{k}=(\varepsilon _{ijk}~S_{mj,i}~\mathbf {e} _{k}\otimes \mathbf {e} _{m})\cdot \mathbf {c} =({\boldsymbol {\nabla }}\times {\boldsymbol {S}})\cdot \mathbf {c} } Therefore, we have ∇ × S = ε i j k S m j , i e k ⊗ e m {\displaystyle {\boldsymbol {\nabla }}\times {\boldsymbol {S}}=\varepsilon _{ijk}~S_{mj,i}~\mathbf {e} _{k}\otimes \mathbf {e} _{m}} </p> <h3>Identities involving the curl of a tensor field</h3> <p>The most commonly used identity involving the curl of a tensor field, T {\displaystyle {\boldsymbol {T}}} , is ∇ × ( ∇ T ) = 0 {\displaystyle {\boldsymbol {\nabla }}\times ({\boldsymbol {\nabla }}{\boldsymbol {T}})={\boldsymbol {0}}} This identity holds for tensor fields of all orders. For the important case of a second-order tensor, S {\displaystyle {\boldsymbol {S}}} , this identity implies that ∇ × ( ∇ S ) = 0 ⟹ S m i , j − S m j , i = 0 {\displaystyle {\boldsymbol {\nabla }}\times ({\boldsymbol {\nabla }}{\boldsymbol {S}})={\boldsymbol {0}}\quad \implies \quad S_{mi,j}-S_{mj,i}=0} </p> <h2 id="derivative-of-the-determinant-of-a-second-order-tensor">Derivative of the determinant of a second-order tensor</h2> <p>The derivative of the determinant of a second order tensor A {\displaystyle {\boldsymbol {A}}} is given by ∂ ∂ A det ( A ) = det ( A ) [ A − 1 ] T . {\displaystyle {\frac {\partial }{\partial {\boldsymbol {A}}}}\det({\boldsymbol {A}})=\det({\boldsymbol {A}})~\left[{\boldsymbol {A}}^{-1}\right]^{\textsf {T}}~.} </p><p>In an orthonormal basis, the components of A {\displaystyle {\boldsymbol {A}}} can be written as a matrix A. In that case, the right hand side corresponds the cofactors of the matrix. </p> Proof <p>Let A {\displaystyle {\boldsymbol {A}}} be a second order tensor and let f ( A ) = det ( A ) {\displaystyle f({\boldsymbol {A}})=\det({\boldsymbol {A}})} . Then, from the definition of the derivative of a scalar valued function of a tensor, we have ∂ f ∂ A : T = d d α det ( A + α T ) | α = 0 = d d α det [ α A ( 1 α I + A − 1 ⋅ T ) ] | α = 0 = d d α [ α 3 det ( A ) det ( 1 α I + A − 1 ⋅ T ) ] | α = 0 . {\displaystyle {\begin{aligned}{\frac {\partial f}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}&=\left.{\cfrac {d}{d\alpha }}\det({\boldsymbol {A}}+\alpha ~{\boldsymbol {T}})\right|_{\alpha =0}\\&=\left.{\cfrac {d}{d\alpha }}\det \left[\alpha ~{\boldsymbol {A}}\left({\cfrac {1}{\alpha }}~{\boldsymbol {\mathit {I}}}+{\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)\right]\right|_{\alpha =0}\\&=\left.{\cfrac {d}{d\alpha }}\left[\alpha ^{3}~\det({\boldsymbol {A}})~\det \left({\cfrac {1}{\alpha }}~{\boldsymbol {\mathit {I}}}+{\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)\right]\right|_{\alpha =0}.\end{aligned}}} </p><p>The determinant of a tensor can be expressed in the form of a characteristic equation in terms of the invariants I 1 , I 2 , I 3 {\displaystyle I_{1},I_{2},I_{3}} using det ( λ I + A ) = λ 3 + I 1 ( A ) λ 2 + I 2 ( A ) λ + I 3 ( A ) . {\displaystyle \det(\lambda ~{\boldsymbol {\mathit {I}}}+{\boldsymbol {A}})=\lambda ^{3}+I_{1}({\boldsymbol {A}})~\lambda ^{2}+I_{2}({\boldsymbol {A}})~\lambda +I_{3}({\boldsymbol {A}}).} </p><p>Using this expansion we can write ∂ f ∂ A : T = d d α [ α 3 det ( A ) ( 1 α 3 + I 1 ( A − 1 ⋅ T ) 1 α 2 + I 2 ( A − 1 ⋅ T ) 1 α + I 3 ( A − 1 ⋅ T ) ) ] | α = 0 = det ( A ) d d α [ 1 + I 1 ( A − 1 ⋅ T ) α + I 2 ( A − 1 ⋅ T ) α 2 + I 3 ( A − 1 ⋅ T ) α 3 ] | α = 0 = det ( A ) [ I 1 ( A − 1 ⋅ T ) + 2 I 2 ( A − 1 ⋅ T ) α + 3 I 3 ( A − 1 ⋅ T ) α 2 ] | α = 0 = det ( A ) I 1 ( A − 1 ⋅ T ) . {\displaystyle {\begin{aligned}{\frac {\partial f}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}&=\left.{\cfrac {d}{d\alpha }}\left[\alpha ^{3}~\det({\boldsymbol {A}})~\left({\cfrac {1}{\alpha ^{3}}}+I_{1}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~{\cfrac {1}{\alpha ^{2}}}+I_{2}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~{\cfrac {1}{\alpha }}+I_{3}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)\right)\right]\right|_{\alpha =0}\\&=\left.\det({\boldsymbol {A}})~{\cfrac {d}{d\alpha }}\left[1+I_{1}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~\alpha +I_{2}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~\alpha ^{2}+I_{3}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~\alpha ^{3}\right]\right|_{\alpha =0}\\&=\left.\det({\boldsymbol {A}})~\left[I_{1}({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}})+2~I_{2}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~\alpha +3~I_{3}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~\alpha ^{2}\right]\right|_{\alpha =0}\\&=\det({\boldsymbol {A}})~I_{1}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)~.\end{aligned}}} </p><p>Recall that the invariant I 1 {\displaystyle I_{1}} is given by I 1 ( A ) = tr A . {\displaystyle I_{1}({\boldsymbol {A}})={\text{tr}}{\boldsymbol {A}}.} </p><p>Hence, ∂ f ∂ A : T = det ( A ) tr ( A − 1 ⋅ T ) = det ( A ) [ A − 1 ] T : T . {\displaystyle {\frac {\partial f}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}=\det({\boldsymbol {A}})~{\text{tr}}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\right)=\det({\boldsymbol {A}})~\left[{\boldsymbol {A}}^{-1}\right]^{\textsf {T}}:{\boldsymbol {T}}.} </p><p>Invoking the arbitrariness of T {\displaystyle {\boldsymbol {T}}} we then have ∂ f ∂ A = det ( A ) [ A − 1 ] T . {\displaystyle {\frac {\partial f}{\partial {\boldsymbol {A}}}}=\det({\boldsymbol {A}})~\left[{\boldsymbol {A}}^{-1}\right]^{\textsf {T}}~.} </p> <h2 id="derivatives-of-the-invariants-of-a-second-order-tensor">Derivatives of the invariants of a second-order tensor</h2> <p>The principal invariants of a second order tensor are I 1 ( A ) = tr A I 2 ( A ) = 1 2 [ ( tr A ) 2 − tr A 2 ] I 3 ( A ) = det ( A ) {\displaystyle {\begin{aligned}I_{1}({\boldsymbol {A}})&={\text{tr}}{\boldsymbol {A}}\\I_{2}({\boldsymbol {A}})&={\tfrac {1}{2}}\left[({\text{tr}}{\boldsymbol {A}})^{2}-{\text{tr}}{{\boldsymbol {A}}^{2}}\right]\\I_{3}({\boldsymbol {A}})&=\det({\boldsymbol {A}})\end{aligned}}} </p><p>The derivatives of these three invariants with respect to A {\displaystyle {\boldsymbol {A}}} are ∂ I 1 ∂ A = 1 ∂ I 2 ∂ A = I 1 1 − A T ∂ I 3 ∂ A = det ( A ) [ A − 1 ] T = I 2 1 − A T ( I 1 1 − A T ) = ( A 2 − I 1 A + I 2 1 ) T {\displaystyle {\begin{aligned}{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}&={\boldsymbol {\mathit {1}}}\\[3pt]{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}&=I_{1}\,{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\\[3pt]{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}&=\det({\boldsymbol {A}})~\left[{\boldsymbol {A}}^{-1}\right]^{\textsf {T}}\\&=I_{2}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}~\left(I_{1}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\right)=\left({\boldsymbol {A}}^{2}-I_{1}~{\boldsymbol {A}}+I_{2}~{\boldsymbol {\mathit {1}}}\right)^{\textsf {T}}\end{aligned}}} </p> Proof <p>From the derivative of the determinant we know that ∂ I 3 ∂ A = det ( A ) [ A − 1 ] T . {\displaystyle {\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}=\det({\boldsymbol {A}})~\left[{\boldsymbol {A}}^{-1}\right]^{\textsf {T}}~.} </p><p>For the derivatives of the other two invariants, let us go back to the characteristic equation det ( λ 1 + A ) = λ 3 + I 1 ( A ) λ 2 + I 2 ( A ) λ + I 3 ( A ) . {\displaystyle \det(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})=\lambda ^{3}+I_{1}({\boldsymbol {A}})~\lambda ^{2}+I_{2}({\boldsymbol {A}})~\lambda +I_{3}({\boldsymbol {A}})~.} </p><p>Using the same approach as for the determinant of a tensor, we can show that ∂ ∂ A det ( λ 1 + A ) = det ( λ 1 + A ) [ ( λ 1 + A ) − 1 ] T . {\displaystyle {\frac {\partial }{\partial {\boldsymbol {A}}}}\det(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})=\det(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})~\left[(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})^{-1}\right]^{\textsf {T}}~.} </p><p>Now the left hand side can be expanded as ∂ ∂ A det ( λ 1 + A ) = ∂ ∂ A [ λ 3 + I 1 ( A ) λ 2 + I 2 ( A ) λ + I 3 ( A ) ] = ∂ I 1 ∂ A λ 2 + ∂ I 2 ∂ A λ + ∂ I 3 ∂ A . {\displaystyle {\begin{aligned}{\frac {\partial }{\partial {\boldsymbol {A}}}}\det(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})&={\frac {\partial }{\partial {\boldsymbol {A}}}}\left[\lambda ^{3}+I_{1}({\boldsymbol {A}})~\lambda ^{2}+I_{2}({\boldsymbol {A}})~\lambda +I_{3}({\boldsymbol {A}})\right]\\&={\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda +{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}~.\end{aligned}}} </p><p>Hence ∂ I 1 ∂ A λ 2 + ∂ I 2 ∂ A λ + ∂ I 3 ∂ A = det ( λ 1 + A ) [ ( λ 1 + A ) − 1 ] T {\displaystyle {\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda +{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}=\det(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})~\left[(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})^{-1}\right]^{\textsf {T}}} or, ( λ 1 + A ) T ⋅ [ ∂ I 1 ∂ A λ 2 + ∂ I 2 ∂ A λ + ∂ I 3 ∂ A ] = det ( λ 1 + A ) 1 . {\displaystyle (\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})^{\textsf {T}}\cdot \left[{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda +{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}\right]=\det(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}})~{\boldsymbol {\mathit {1}}}~.} </p><p>Expanding the right hand side and separating terms on the left hand side gives ( λ 1 + A T ) ⋅ [ ∂ I 1 ∂ A λ 2 + ∂ I 2 ∂ A λ + ∂ I 3 ∂ A ] = [ λ 3 + I 1 λ 2 + I 2 λ + I 3 ] 1 {\displaystyle \left(\lambda ~{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}}^{\textsf {T}}\right)\cdot \left[{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda +{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}\right]=\left[\lambda ^{3}+I_{1}~\lambda ^{2}+I_{2}~\lambda +I_{3}\right]{\boldsymbol {\mathit {1}}}} </p><p>or, [ ∂ I 1 ∂ A λ 3 + ∂ I 2 ∂ A λ 2 + ∂ I 3 ∂ A λ ] 1 + A T ⋅ ∂ I 1 ∂ A λ 2 + A T ⋅ ∂ I 2 ∂ A λ + A T ⋅ ∂ I 3 ∂ A = [ λ 3 + I 1 λ 2 + I 2 λ + I 3 ] 1 . {\displaystyle {\begin{aligned}\left[{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{3}\right.&\left.+{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}~\lambda \right]{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda +{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}\\&=\left[\lambda ^{3}+I_{1}~\lambda ^{2}+I_{2}~\lambda +I_{3}\right]{\boldsymbol {\mathit {1}}}~.\end{aligned}}} </p><p>If we define I 0 := 1 {\displaystyle I_{0}:=1} and I 4 := 0 {\displaystyle I_{4}:=0} , we can write the above as [ ∂ I 1 ∂ A λ 3 + ∂ I 2 ∂ A λ 2 + ∂ I 3 ∂ A λ + ∂ I 4 ∂ A ] 1 + A T ⋅ ∂ I 0 ∂ A λ 3 + A T ⋅ ∂ I 1 ∂ A λ 2 + A T ⋅ ∂ I 2 ∂ A λ + A T ⋅ ∂ I 3 ∂ A = [ I 0 λ 3 + I 1 λ 2 + I 2 λ + I 3 ] 1 . {\displaystyle {\begin{aligned}\left[{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{3}\right.&\left.+{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}~\lambda +{\frac {\partial I_{4}}{\partial {\boldsymbol {A}}}}\right]{\boldsymbol {\mathit {1}}}+{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{0}}{\partial {\boldsymbol {A}}}}~\lambda ^{3}+{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~\lambda ^{2}+{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~\lambda +{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}\\&=\left[I_{0}~\lambda ^{3}+I_{1}~\lambda ^{2}+I_{2}~\lambda +I_{3}\right]{\boldsymbol {\mathit {1}}}~.\end{aligned}}} </p><p>Collecting terms containing various powers of λ, we get λ 3 ( I 0 1 − ∂ I 1 ∂ A 1 − A T ⋅ ∂ I 0 ∂ A ) + λ 2 ( I 1 1 − ∂ I 2 ∂ A 1 − A T ⋅ ∂ I 1 ∂ A ) + λ ( I 2 1 − ∂ I 3 ∂ A 1 − A T ⋅ ∂ I 2 ∂ A ) + ( I 3 1 − ∂ I 4 ∂ A 1 − A T ⋅ ∂ I 3 ∂ A ) = 0 . {\displaystyle {\begin{aligned}\lambda ^{3}&\left(I_{0}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{0}}{\partial {\boldsymbol {A}}}}\right)+\lambda ^{2}\left(I_{1}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}\right)+\\&\qquad \qquad \lambda \left(I_{2}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}\right)+\left(I_{3}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{4}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}\right)=0~.\end{aligned}}} </p><p>Then, invoking the arbitrariness of λ, we have I 0 1 − ∂ I 1 ∂ A 1 − A T ⋅ ∂ I 0 ∂ A = 0 I 1 1 − ∂ I 2 ∂ A 1 − I 2 1 − ∂ I 3 ∂ A 1 − A T ⋅ ∂ I 2 ∂ A = 0 I 3 1 − ∂ I 4 ∂ A 1 − A T ⋅ ∂ I 3 ∂ A = 0 . {\displaystyle {\begin{aligned}I_{0}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{0}}{\partial {\boldsymbol {A}}}}&=0\\I_{1}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-I_{2}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}&=0\\I_{3}~{\boldsymbol {\mathit {1}}}-{\frac {\partial I_{4}}{\partial {\boldsymbol {A}}}}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\cdot {\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}&=0~.\end{aligned}}} </p><p>This implies that ∂ I 1 ∂ A = 1 ∂ I 2 ∂ A = I 1 1 − A T ∂ I 3 ∂ A = I 2 1 − A T ( I 1 1 − A T ) = ( A 2 − I 1 A + I 2 1 ) T {\displaystyle {\begin{aligned}{\frac {\partial I_{1}}{\partial {\boldsymbol {A}}}}&={\boldsymbol {\mathit {1}}}\\{\frac {\partial I_{2}}{\partial {\boldsymbol {A}}}}&=I_{1}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\\{\frac {\partial I_{3}}{\partial {\boldsymbol {A}}}}&=I_{2}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}~\left(I_{1}~{\boldsymbol {\mathit {1}}}-{\boldsymbol {A}}^{\textsf {T}}\right)=\left({\boldsymbol {A}}^{2}-I_{1}~{\boldsymbol {A}}+I_{2}~{\boldsymbol {\mathit {1}}}\right)^{\textsf {T}}\end{aligned}}} </p> <h2 id="derivative-of-the-second-order-identity-tensor">Derivative of the second-order identity tensor</h2> <p>Let 1 {\displaystyle {\boldsymbol {\mathit {1}}}} be the second order identity tensor. Then the derivative of this tensor with respect to a second order tensor A {\displaystyle {\boldsymbol {A}}} is given by ∂ 1 ∂ A : T = 0 : T = 0 {\displaystyle {\frac {\partial {\boldsymbol {\mathit {1}}}}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}={\boldsymbol {\mathsf {0}}}:{\boldsymbol {T}}={\boldsymbol {\mathit {0}}}} This is because 1 {\displaystyle {\boldsymbol {\mathit {1}}}} is independent of A {\displaystyle {\boldsymbol {A}}} . </p> <h2 id="derivative-of-a-second-order-tensor-with-respect-to-itself">Derivative of a second-order tensor with respect to itself</h2> <p>Let A {\displaystyle {\boldsymbol {A}}} be a second order tensor. Then ∂ A ∂ A : T = [ ∂ ∂ α ( A + α T ) ] α = 0 = T = I : T {\displaystyle {\frac {\partial {\boldsymbol {A}}}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}=\left[{\frac {\partial }{\partial \alpha }}({\boldsymbol {A}}+\alpha ~{\boldsymbol {T}})\right]_{\alpha =0}={\boldsymbol {T}}={\boldsymbol {\mathsf {I}}}:{\boldsymbol {T}}} </p><p>Therefore, ∂ A ∂ A = I {\displaystyle {\frac {\partial {\boldsymbol {A}}}{\partial {\boldsymbol {A}}}}={\boldsymbol {\mathsf {I}}}} </p><p>Here I {\displaystyle {\boldsymbol {\mathsf {I}}}} is the fourth order identity tensor. In index notation with respect to an orthonormal basis I = δ i k δ j l e i ⊗ e j ⊗ e k ⊗ e l {\displaystyle {\boldsymbol {\mathsf {I}}}=\delta _{ik}~\delta _{jl}~\mathbf {e} _{i}\otimes \mathbf {e} _{j}\otimes \mathbf {e} _{k}\otimes \mathbf {e} _{l}} </p><p>This result implies that ∂ A T ∂ A : T = I T : T = T T {\displaystyle {\frac {\partial {\boldsymbol {A}}^{\textsf {T}}}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}={\boldsymbol {\mathsf {I}}}^{\textsf {T}}:{\boldsymbol {T}}={\boldsymbol {T}}^{\textsf {T}}} where I T = δ j k δ i l e i ⊗ e j ⊗ e k ⊗ e l {\displaystyle {\boldsymbol {\mathsf {I}}}^{\textsf {T}}=\delta _{jk}~\delta _{il}~\mathbf {e} _{i}\otimes \mathbf {e} _{j}\otimes \mathbf {e} _{k}\otimes \mathbf {e} _{l}} </p><p>Therefore, if the tensor A {\displaystyle {\boldsymbol {A}}} is symmetric, then the derivative is also symmetric and we get ∂ A ∂ A = I ( s ) = 1 2 ( I + I T ) {\displaystyle {\frac {\partial {\boldsymbol {A}}}{\partial {\boldsymbol {A}}}}={\boldsymbol {\mathsf {I}}}^{(s)}={\frac {1}{2}}~\left({\boldsymbol {\mathsf {I}}}+{\boldsymbol {\mathsf {I}}}^{\textsf {T}}\right)} where the symmetric fourth order identity tensor is I ( s ) = 1 2 ( δ i k δ j l + δ i l δ j k ) e i ⊗ e j ⊗ e k ⊗ e l {\displaystyle {\boldsymbol {\mathsf {I}}}^{(s)}={\frac {1}{2}}~(\delta _{ik}~\delta _{jl}+\delta _{il}~\delta _{jk})~\mathbf {e} _{i}\otimes \mathbf {e} _{j}\otimes \mathbf {e} _{k}\otimes \mathbf {e} _{l}} </p> <h2 id="derivative-of-the-inverse-of-a-second-order-tensor">Derivative of the inverse of a second-order tensor</h2> <p>Let A {\displaystyle {\boldsymbol {A}}} and T {\displaystyle {\boldsymbol {T}}} be two second order tensors, then ∂ ∂ A ( A − 1 ) : T = − A − 1 ⋅ T ⋅ A − 1 {\displaystyle {\frac {\partial }{\partial {\boldsymbol {A}}}}\left({\boldsymbol {A}}^{-1}\right):{\boldsymbol {T}}=-{\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\cdot {\boldsymbol {A}}^{-1}} In index notation with respect to an orthonormal basis ∂ A i j − 1 ∂ A k l T k l = − A i k − 1 T k l A l j − 1 ⟹ ∂ A i j − 1 ∂ A k l = − A i k − 1 A l j − 1 {\displaystyle {\frac {\partial A_{ij}^{-1}}{\partial A_{kl}}}~T_{kl}=-A_{ik}^{-1}~T_{kl}~A_{lj}^{-1}\implies {\frac {\partial A_{ij}^{-1}}{\partial A_{kl}}}=-A_{ik}^{-1}~A_{lj}^{-1}} We also have ∂ ∂ A ( A − T ) : T = − A − T ⋅ T T ⋅ A − T {\displaystyle {\frac {\partial }{\partial {\boldsymbol {A}}}}\left({\boldsymbol {A}}^{-{\textsf {T}}}\right):{\boldsymbol {T}}=-{\boldsymbol {A}}^{-{\textsf {T}}}\cdot {\boldsymbol {T}}^{\textsf {T}}\cdot {\boldsymbol {A}}^{-{\textsf {T}}}} In index notation ∂ A j i − 1 ∂ A k l T k l = − A j k − 1 T l k A l i − 1 ⟹ ∂ A j i − 1 ∂ A k l = − A l i − 1 A j k − 1 {\displaystyle {\frac {\partial A_{ji}^{-1}}{\partial A_{kl}}}~T_{kl}=-A_{jk}^{-1}~T_{lk}~A_{li}^{-1}\implies {\frac {\partial A_{ji}^{-1}}{\partial A_{kl}}}=-A_{li}^{-1}~A_{jk}^{-1}} If the tensor A {\displaystyle {\boldsymbol {A}}} is symmetric then ∂ A i j − 1 ∂ A k l = − 1 2 ( A i k − 1 A j l − 1 + A i l − 1 A j k − 1 ) {\displaystyle {\frac {\partial A_{ij}^{-1}}{\partial A_{kl}}}=-{\cfrac {1}{2}}\left(A_{ik}^{-1}~A_{jl}^{-1}+A_{il}^{-1}~A_{jk}^{-1}\right)} </p> Proof <p>Recall that ∂ 1 ∂ A : T = 0 {\displaystyle {\frac {\partial {\boldsymbol {\mathit {1}}}}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}={\boldsymbol {\mathit {0}}}} </p><p>Since A − 1 ⋅ A = 1 {\displaystyle {\boldsymbol {A}}^{-1}\cdot {\boldsymbol {A}}={\boldsymbol {\mathit {1}}}} , we can write ∂ ∂ A ( A − 1 ⋅ A ) : T = 0 {\displaystyle {\frac {\partial }{\partial {\boldsymbol {A}}}}\left({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {A}}\right):{\boldsymbol {T}}={\boldsymbol {\mathit {0}}}} </p><p>Using the product rule for second order tensors ∂ ∂ S [ F 1 ( S ) ⋅ F 2 ( S ) ] : T = ( ∂ F 1 ∂ S : T ) ⋅ F 2 + F 1 ⋅ ( ∂ F 2 ∂ S : T ) {\displaystyle {\frac {\partial }{\partial {\boldsymbol {S}}}}[{\boldsymbol {F}}_{1}({\boldsymbol {S}})\cdot {\boldsymbol {F}}_{2}({\boldsymbol {S}})]:{\boldsymbol {T}}=\left({\frac {\partial {\boldsymbol {F}}_{1}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)\cdot {\boldsymbol {F}}_{2}+{\boldsymbol {F}}_{1}\cdot \left({\frac {\partial {\boldsymbol {F}}_{2}}{\partial {\boldsymbol {S}}}}:{\boldsymbol {T}}\right)} </p><p>we get ∂ ∂ A ( A − 1 ⋅ A ) : T = ( ∂ A − 1 ∂ A : T ) ⋅ A + A − 1 ⋅ ( ∂ A ∂ A : T ) = 0 {\displaystyle {\frac {\partial }{\partial {\boldsymbol {A}}}}({\boldsymbol {A}}^{-1}\cdot {\boldsymbol {A}}):{\boldsymbol {T}}=\left({\frac {\partial {\boldsymbol {A}}^{-1}}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}\right)\cdot {\boldsymbol {A}}+{\boldsymbol {A}}^{-1}\cdot \left({\frac {\partial {\boldsymbol {A}}}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}\right)={\boldsymbol {\mathit {0}}}} or, ( ∂ A − 1 ∂ A : T ) ⋅ A = − A − 1 ⋅ T {\displaystyle \left({\frac {\partial {\boldsymbol {A}}^{-1}}{\partial {\boldsymbol {A}}}}:{\boldsymbol {T}}\right)\cdot {\boldsymbol {A}}=-{\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}} </p><p>Therefore, ∂ ∂ A ( A − 1 ) : T = − A − 1 ⋅ T ⋅ A − 1 {\displaystyle {\frac {\partial }{\partial {\boldsymbol {A}}}}\left({\boldsymbol {A}}^{-1}\right):{\boldsymbol {T}}=-{\boldsymbol {A}}^{-1}\cdot {\boldsymbol {T}}\cdot {\boldsymbol {A}}^{-1}} </p> <h2 id="integration-by-parts">Integration by parts</h2> <p>Another important operation related to tensor derivatives in continuum mechanics is integration by parts. The formula for integration by parts can be written as ∫ Ω F ⊗ ∇ G d Ω = ∫ Γ n ⊗ ( F ⊗ G ) d Γ − ∫ Ω G ⊗ ∇ F d Ω {\displaystyle \int _{\Omega }{\boldsymbol {F}}\otimes {\boldsymbol {\nabla }}{\boldsymbol {G}}\,d\Omega =\int _{\Gamma }\mathbf {n} \otimes ({\boldsymbol {F}}\otimes {\boldsymbol {G}})\,d\Gamma -\int _{\Omega }{\boldsymbol {G}}\otimes {\boldsymbol {\nabla }}{\boldsymbol {F}}\,d\Omega } </p><p>where F {\displaystyle {\boldsymbol {F}}} and G {\displaystyle {\boldsymbol {G}}} are differentiable tensor fields of arbitrary order, n {\displaystyle \mathbf {n} } is the unit outward normal to the domain over which the tensor fields are defined, ⊗ {\displaystyle \otimes } represents a generalized tensor product operator, and ∇ {\displaystyle {\boldsymbol {\nabla }}} is a generalized gradient operator. When F {\displaystyle {\boldsymbol {F}}} is equal to the identity tensor, we get the <a href="/facts/Divergence_theorem/pUGKF1N2">divergence theorem</a> ∫ Ω ∇ G d Ω = ∫ Γ n ⊗ G d Γ . {\displaystyle \int _{\Omega }{\boldsymbol {\nabla }}{\boldsymbol {G}}\,d\Omega =\int _{\Gamma }\mathbf {n} \otimes {\boldsymbol {G}}\,d\Gamma \,.} </p><p>We can express the formula for integration by parts in Cartesian index notation as ∫ Ω F i j k . . . . G l m n . . . , p d Ω = ∫ Γ n p F i j k . . . G l m n . . . d Γ − ∫ Ω G l m n . . . F i j k . . . , p d Ω . {\displaystyle \int _{\Omega }F_{ijk....}\,G_{lmn...,p}\,d\Omega =\int _{\Gamma }n_{p}\,F_{ijk...}\,G_{lmn...}\,d\Gamma -\int _{\Omega }G_{lmn...}\,F_{ijk...,p}\,d\Omega \,.} </p><p>For the special case where the tensor product operation is a contraction of one index and the gradient operation is a divergence, and both F {\displaystyle {\boldsymbol {F}}} and G {\displaystyle {\boldsymbol {G}}} are second order tensors, we have ∫ Ω F ⋅ ( ∇ ⋅ G ) d Ω = ∫ Γ n ⋅ ( G ⋅ F T ) d Γ − ∫ Ω ( ∇ F ) : G T d Ω . {\displaystyle \int _{\Omega }{\boldsymbol {F}}\cdot ({\boldsymbol {\nabla }}\cdot {\boldsymbol {G}})\,d\Omega =\int _{\Gamma }\mathbf {n} \cdot \left({\boldsymbol {G}}\cdot {\boldsymbol {F}}^{\textsf {T}}\right)\,d\Gamma -\int _{\Omega }({\boldsymbol {\nabla }}{\boldsymbol {F}}):{\boldsymbol {G}}^{\textsf {T}}\,d\Omega \,.} </p><p>In index notation, ∫ Ω F i j G p j , p d Ω = ∫ Γ n p F i j G p j d Γ − ∫ Ω G p j F i j , p d Ω . {\displaystyle \int _{\Omega }F_{ij}\,G_{pj,p}\,d\Omega =\int _{\Gamma }n_{p}\,F_{ij}\,G_{pj}\,d\Gamma -\int _{\Omega }G_{pj}\,F_{ij,p}\,d\Omega \,.} </p> <h2 id="see-also">See also</h2> <ul><li><a href="/facts/Covariant_derivative/sRGN0Xh6">Covariant derivative</a></li> <li><a href="/facts/Ricci_calculus/We4yudeq">Ricci calculus</a></li></ul> <h2 id="references">References</h2> <ol> <li id="fn:1"><p>Simo, J. C.; Hughes, T. J. R. (1998). Computational Inelasticity. Springer. doi:10.1007/b98904. ISBN 978-0-387-97520-7. <a href="978-0-387-97520-7" target="_blank">978-0-387-97520-7</a> <a href="#fnref:1" class="footnote-back-ref">↩</a></p></li> <li id="fn:2"><p>Marsden, Jerrold E.; Hughes, Thomas J. R. (2000). Mathematical Foundations of Elasticity. Dover. ISBN 978-0-486-678658. <a href="978-0-486-678658" target="_blank">978-0-486-678658</a> <a href="#fnref:2" class="footnote-back-ref">↩</a></p></li> <li id="fn:3"><p>Ogden, R. W. (2000). Nonlinear Elastic Deformations. Dover. ISBN 978-0-486-696485. <a href="978-0-486-696485" target="_blank">978-0-486-696485</a> <a href="#fnref:3" class="footnote-back-ref">↩</a></p></li> <li id="fn:4"><p>Hjelmstad, Keith (2004). Fundamentals of Structural Mechanics. Springer Science & Business Media. p. 45. ISBN 978-0-387-233307. <a href="978-0-387-233307" target="_blank">978-0-387-233307</a> <a href="#fnref:4" class="footnote-back-ref">↩</a></p></li> <li id="fn:5"><p>Hjelmstad, Keith (2004). Fundamentals of Structural Mechanics. Springer Science & Business Media. p. 45. ISBN 978-0-387-233307. <a href="978-0-387-233307" target="_blank">978-0-387-233307</a> <a href="#fnref:5" class="footnote-back-ref">↩</a></p></li> </ol>