Dear all,

I think I ran into a minor mistake in the pair_gran documentation page

(lammps-7Mar10), where it is written that the shear modulus G and the Young's

modulus E are related through the formula :

G = E / (1(1+nu)),

whereas in several textbooks I found the following relation:

G = E / (2(1+nu)).

Did I missed something? Or could you make the corresponding correction?

Best regards,

Laurent Joly

I'm fairly certain the doc page represents what is actually

implemented in the code. It may be there is a factor of 2 difference

in the definition of some constant, but you should be able to compensate

by altering your input params by the same factor of 2. I am CCing

Jeremy Lechman here at Sandia who knows more about the

mapping to G,E than I do - to see which version of the formula

below he agrees with.

Steve

Thank you for your answer. In fact this formula is just given in the doc to

explain how to map the elastic constants Kn and Kt to the Young's modulus of

the material, but it is not implemented in the code, so it is clearly not a

big problem. It may not have been detected until now since most people seems

to work in reduced units with more or less the same parameters Kn and Kt than

in the example problem...

Best regards,

Laurent Joly

Hi Laurent,

Thanks for pointing this out. The correct formula for relating shear modulus to Young's modulus and Poisson ratio is:

G = E / (2(1+nu)).

I believe this is "just" a typo in the documentation page, since the user actually specifies the spring constants Kn and Kt which are reduced parameters from elastic material properties. Hopefully this hasn't lead to too much confusion.

Best,

Jeremy

Fixed it on the doc page - thanks

Steve