# Internal Energy equation in viscous compressible floe equations

 Register Blogs Members List Search Today's Posts Mark Forums Read

 August 17, 2015, 15:34 Internal Energy equation in viscous compressible floe equations #1 Member   Mihir Makwana Join Date: May 2015 Posts: 79 Rep Power: 10 How can I use FVM to discretize the term sigma : D in the internal energy equation http://i.imgur.com/3RaYm5p.jpg where sigma and D are given by http://i.imgur.com/6K65Lih.jpg I am not able to expand the term sigma : D Please help. Thanks in Advance - Mihir

 August 17, 2015, 17:03 #2 Senior Member   Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,599 Rep Power: 70 If you want using a FVM, you have to recast the equation in the divergence form and integrate over a finite volume each term. Using the expression of sigma, the term D:D you get is a scalar function that appears as volume source

August 18, 2015, 01:21
#3
Member

Mihir Makwana
Join Date: May 2015
Posts: 79
Rep Power: 10
Quote:
 Originally Posted by FMDenaro If you want using a FVM, you have to recast the equation in the divergence form and integrate over a finite volume each term. Using the expression of sigma, the term D:D you get is a scalar function that appears as volume source
sir, if i use tensors then sigma : D gives

http://i.imgur.com/UmzDCrI.jpg

How do i convert this equation to divergence form ?

 August 18, 2015, 02:45 #4 Member   Mihir Makwana Join Date: May 2015 Posts: 79 Rep Power: 10 sir say i split D then one of the term is http://i.imgur.com/zkbee0D.jpg here Tau is sigma so the 1st term on r.h.s is in divergence form but the second is not

 August 18, 2015, 03:13 #5 Senior Member   Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,599 Rep Power: 70 no, you do not have all terms in te internal energy equation in divergence form...that respects the physical fact that such energy form does not have a conservative formulation. The term D: D is a pointwise source term, you have to integrate over the local volume in a FVM and discretize the volume integral. If you want to work with a fully divergence form then you have to adopt the total energy equation

August 18, 2015, 04:58
#6
Member

Mihir Makwana
Join Date: May 2015
Posts: 79
Rep Power: 10
Quote:
 Originally Posted by FMDenaro no, you do not have all terms in te internal energy equation in divergence form...that respects the physical fact that such energy form does not have a conservative formulation. The term D: D is a pointwise source term, you have to integrate over the local volume in a FVM and discretize the volume integral. If you want to work with a fully divergence form then you have to adopt the total energy equation
ok.

1) As i am using FVM, i cannot integrate the second term in RHS of

http://i.imgur.com/zkbee0D.jpg

Right ??

or is there a way I can integrate it over the C.V

2) what is D: D ?

 August 18, 2015, 05:02 #7 Senior Member   Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,599 Rep Power: 70 you can always integrate each term of the equation over a finite volume...when the term is in divergence form you can apply Gauss and get the surface integral of the fluxes, conversely the integral to discretize remains the volume integral (see for example the book of Peric & Ferziger). D: D is nothing else that the double dot product between the symmetric velocity gradient

August 18, 2015, 05:04
#8
Member

Mihir Makwana
Join Date: May 2015
Posts: 79
Rep Power: 10
Quote:
 Originally Posted by FMDenaro you can always integrate each term of the equation over a finite volume...when the term is in divergence form you can apply Gauss and get the surface integral of the fluxes, conversely the integral to discretize remains the volume integral (see for example the book of Peric & Ferziger). D: D is nothing else that the double dot product between the symmetric velocity gradient
1) ok.

2) but i need to find sigma : D and not D : D

 August 18, 2015, 05:11 #9 Senior Member   Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,599 Rep Power: 70 sigma: D is nothing else that a product by an isotropic tensor (I: D) added with D: D for brevity I disregarded the coefficients