*INITIAL_SOLID_VOLUME Purpose: Recalculate and reset initial volume of solid elements using material models with EOS before analysis if the original nodal position has been moved by nodal projections in contact initialization. This option eliminates calculation of non-physical initial hydrostatic pressure due to the nodal repositioning. Author: Jason Wang Date: 2018-07-09 08:36:00 -0700 (Mon, 09 Jul 2018) New Revision: 10169 Added: trunk/Vol_I_Keyword/INITIAL_SOLID_VOLUME.docx Modified: trunk/Vol_I_Keyword/file_list.txt Log: Add new keyword *initial_solid_volume to avoid unexpected hydrostatic pressure created by TIED contact [in materials that have an EOS] Bugzilla 13713 indirectly led to the development of INITIAL_SOLID_VOLUME __________________________________________________________ Q: How would I go about figuring out which eos models work with mat 9 for solids? ANS1 from jd: Interesting question. I don't have a definite answer or know what the code will actually allow, but mat 9 is for fluids so you would only use an EOS that is appropriate for fluids (gas or liquid). It's usually paired with EOS 1 or 4. You certainly wouldn't pair it with an EOS intended for high explosives (e.g., JWL, JWLB) or for soils (COMPACTION-type EOS's). I can't recall encountering a situation where the code wrote a warning or error message complaining about a disallowed pairing of a particular EOS with a particular material that needs an EOS. For example, consider the attached detcord.k. If you switch things around so part 1 uses eosid 2 and part 2 uses eosid 1, the run will fail after about 100 cycles, but there is no input trap that's triggered. ANS2 from db: jd is right. There is no check for compatibility between the EOS and the material model. I don't think I'd attempt to document which are reasonable pairs of EOS and material models. It isn't all that clear cut. For example, you could use the ignition and growth EOS with mat 9, even though it is an explosive EOS, because it doesn't need the programmed burn that mat 8 supplies. There are a lot of feasible combinations (but not necessarily recommended combinations unless you really know what you are doing) and putting all of the IF-tests in the documentation would be a major task. 4/12/19 ____________________________________________________________________ When modeling with continuum elements, certain material models will require that an equation-of-state (EOS) be defined. This is done using the keyword *EOS_option and the variable EOSID in *PART. For details on when an EOS is required, see the MATERIAL MODEL REFERENCE TABLES in the beginning of the *MAT section of the User's Manual. In that table, there's a column labeled EOS. If a "Y" appears in that column, then the corresponding material model requires an equation-of-state when used with solid elements, shell formulations 13/14/15, or tshell formulations 3/5/7. For more information on equations-of-state, see the *EOS section of the User's Manual, http://www.dynasupport.com/howtos/general/equation-of-state, and p. 10.6 in Du Bois's crash notesi (no longer available). In some situations, an EOS is required in order to accurately simulate material behavior. An EOS determines the hydrostatic, or bulk, behavior of the material by calculating pressure as a function of density and perhaps, energy and/or temperature. Situations that call for an EOS are characterized by very high strain rates, material pressures far in excess of yield stress, and propagation of shock waves. Of course, these phenomena are very much interrelated. EOS_LINEAR_POLYNOMIAL or EOS_GRUNEISEN are probably the most commonly used EOS forms for non-gaseous materials. Gruneisen parameters are available for many materials including metals. Total stress is the sum of deviatoric stress and pressure. The mean stress (sig1 + sig2 + sig3)/3 is equal to the pressure. Constitutive models which do NOT employ an EOS calculate total stress directly. In these models, the pressure component of total stress is based only on volumetric strain. For instance, for an elastic material, p = K * mu where K is the bulk modulus and mu = rho/rho0 - 1. Material models that require an accompanying EOS calculate only the deviatoric component of stress, i.e. the strength behavior, whereas the EOS calculates the pressure component of total stress, i.e., the hydrostatic behavior. If you're using a material model where an EOS is required, you can achieve simple bulk behavior by using *eos_linear_polynomial and setting C1 to the bulk modulus = E/(3 * (1-2*PR)) and all the other C terms to zero. I would only recommend this approach if strain rates are low to moderate. Strain rates in an auto crash would qualify as moderate. The user has the option of providing a user-defined subroutine to describe an equation-of-state. A template of such user-defined subroutines is included in the file dyn21b.f. See Appendix B in the Users Manual and the command *EOS_USER_DEFINED. The book "High Velocity Impact Dynamics", edited by Zukas (1990, John Wiley and Sons) is a good reference on the subject of material behavior at high strain rates. EOS parameters for approximately 50 materials are given in "Equation of State and Strength Properties of Selected Materials", Daniel J. Steinberg, Lawrence Livermore National Laboratory, 1991 (Change 1 issued 1996), UCRL-MA-106439. LLNL prohibits this work from being posted electronically, but in an email to jpd from Stephanie Black, LLNL, 8/31/2011, 8:14 am, it can be shared with customers in "hard copy only". See notes regarding distribution from Genevieve Silveroli in notes file "steinberg_INTERNAL" (2022). -jday Regarding EOS_TABULATED_COMPACTION and EOS_TABULATED: The manual isn't very specific. The notes I have indicate the following: - The eVi terms (abscissa of the curve) represent ln(relative volume) and thus are negative in compression. - eVi = ln(relative volume) values should be given in descending order, that is, tensile (positive) value first and largest compression (most negative) value last. - Pressure is positive in compression. If gamma = 0, Ci is equal to pressure on the loading curve. Thus Ci should have an algebraic sign opposite of eVi. When there's an EOS, the initial stress values given in *initial_stress are adjusted so that the mean stress (pressure) is in agreement with the EOS. In other words, -(SIGXX + SIGYY + SIGZZ)/3 - stress adjustment = initial pressure from EOS or stress adjustment = -(SIGXX + SIGYY + SIGZZ)/3 - initial pressure from EOS actual initial sigxx = SIGXX + stress adjustment actual initial sigyy = SIGYY + stress adjustment actual initial sigzz = SIGZZ + stress adjustment See http://ftp.lstc.com/anonymous/outgoing/support/FAQ_kw/mat16.initstress.k ______________________________________________________________________ The following comes from James M. Kennedy KBS2 Inc. April 23, 2010 An excellent reference book you might consider: Zukas, J.A., "Intoduction to Hydrocodes", Studies in Applied Mechanics, Vol. 49, Elsevier, 2004. http://www.elsevier.com/wps/find/bookdescription.authors/701885/description#description Perhaps the following report and paper may have some information you desire: Davydov, B.I., "Equation of State for Solid Bodies", Report AD0600614, Foreign Techology Division Air Force Systems Command Wright-Patterson AFB Ohio, March, 1964. http://oai.dtic.mil/oai/oai?verb=getRecord&metadataPrefix=html&identifier=AD0600614 http://www.ntis.gov/search/product.aspx?ABBR=AD600614 Men''shikov, G.P., "An Equation of State for Solids at High Pressure", Combustion, Explosion, and Shock Waves, Vol. 17, No. 2, pp. 215-222, March, 1981. http://www.springerlink.com/content/xn701j6277t72h30/ A review article on Gruneisen equation of state Mendoza, E., "The Equation of State for Solids 1843-1926", European Journal of Physics, Vol. 3, pp. 181-187, 1982. http://www.iop.org/EJ/abstract/0143-0807/3/3/010 An abstract that might be helpful: http://gsa.confex.com/gsa/2003AM/finalprogram/abstract_60195.htm An introductory note: http://www.ccl.net/cca/documents/dyoung/topics-orig/eq_state.html