14th U.S. National Congress on Computational Mechanics July 17-20, 2017, Montreal, Quebec, Canada Title: Multi-Physics Modeling of Thermomechanical Response of Ultrasonically Activated Soft Tissue Author(s): *Rahul , Suvranu De, Rensselaer Polytechnic Inst.. Ultrasonic surgical instruments have been gaining popularity among surgeons in the last decade. An increasing number of surgical procedures including but not limited to head, neck, gynecological, colorectal and gastrointestinal surgeries are performed using ultrasonic surgical instruments [1]. These instruments utilize ultrasonic vibrations to cut, coagulate and dissect tissues, and seal vessels. They have been proven to be superior to conventional instruments and techniques such as electrosurgical and laser-based devices as they impose lesser thermal injury, desiccation and charring, lower mean blood loss during surgery, no risk of stray current, neuromuscular stimulation, lesser operation time and post-operative pain, and no smoke during the operation to occlude laparoscopic view [2]. Despite the increasing popularity of ultrasound-based surgical procedures, the affects of cellular level mechanisms on the thermomechanical response of ultrasonically activated soft tissues have not been understood completely. We have developed a multi-physics model to investigate the effects of cavitation, due to large transient pressure changes, on the thermomechanical response of soft tissue subjected to ultrasound vibrations [3]. Cellular level cavitation effects have been incorporated in the tissue level continuum model to accurately determine the thermodynamic states such as temperature and pressure. A viscoelastic material model is assumed for the macromechanical response of the tissue. The cavitation model based equation-of-state provides the additional pressure arising from evaporation of intracellular and cellular water by absorbing heat due to viscoelastic heating in the tissue, and temperature to the continuum level thermomechanical model. The thermomechanical response of soft tissue is studied for the operational range of frequencies of oscillations and applied loads for typical ultrasonically activated surgical instruments. The model is shown to capture characteristics of ultrasonically activated soft tissue deformation and temperature evolution. At the cellular level, evaporation of water below the boiling temperature under ambient conditions is indicative of protein denaturation around the temperature threshold for coagulation of tissues. Further, with increasing operating frequency (or loading), the temperature rises faster leading to rapid evaporation of tissue cavity water, which may lead to accelerated protein denaturation and coagulation. References: [1] Sankaranarayanan, G., Resapu, R.R., Jones, D.B., et al. (2013) Surg. Endosc. 27, 3056-3072. [2] Amaral, J. F., Chrostek, C. A. (1997) Minim. Invasive Ther. Allied Technol. 6, 324-331. [3] Rahul, De, S., (2017) Comput. Methods Appl. Mech. Eng. 314, 71-84. 14th U.S. National Congress on Computational Mechanics July 17-20, 2017, Montreal, Quebec, Canada Title: Plastic Deformation and Non-linear Damage Evolution Modeling for Pipeline Steel Author(s): *Hari Simha , canmetMaterials, NRCAN. Field deformation data for X70 pipeline steel, obtained using digital-image correlation, is used to calibrate both plasticity and damage models. Experimental deformation fields were obtained on uni-axially loaded round bars, notched round bars, notched flat plates, and plates with a central hole. The well-known Hill48 model is used to describe yield and anisotropy of strength. Using virtual extensometers that are strategically located, anisotropy coefficients are calibrated using the DIC data. Implicit finite element simulations using the Abaqus finite element software are used for the latter effort. Onset of localization and subsequent damage evolution is modeled using the frameworks developed by Wierzbicki and co-workers, [1, 2]. A strain-based failure surface that depends on both the Lode angle (third invariant of the stress tensor) and the tri-axiality (second invariant of the stress tensor) is used as the threshold for localization. The novelty in the damage model is the use of an evolution law that is reminiscent of the flow curves used in unified plasticity; specifically, the Bodner-Partom model, [3]. On account of this model, low values of damage exists at all load levels, but starts to grow significantly only when the localization threshold described by the failure surface is exceeded. This feature circumvents the need to assume damage levels greater than unity to model crack propagation. The damage-mechanics framework is implemented in Abaqus/explicit. To mitigate mesh dependence owing to softening of the strength, a non-local approach is used. Examples of ductile damage propagation in compact and single edged notch tension samples are shown. Particularly, the importance of the Lode angle in modeling the flat-to-slant transition in crack propagation is emphasized. Where possible, fracture mechanics experiments are modeled using the developed framework. [1] Bai, Y. and Wierzbicki, T., 2008. A new model of metal plasticity and fracture with pressure and Lode dependence. International journal of plasticity, 24(6), pp.1071-1096. [2] Xue, L. and Wierzbicki, T., 2008. Ductile fracture initiation and propagation modeling using damage plasticity theory. Engineering Fracture Mechanics, 75(11), pp.3276-3293. [3] Bodner, S. and Partom, Y., 1975. Constitutive equations for elastic-viscoplastic strain-hardening materials. Journal of Applied Mechanics, 42(2), pp.385-389. 14th U.S. National Congress on Computational Mechanics July 17-20, 2017, Montreal, Quebec, Canada Title: An Equation-Free Multiscale Method: Extension of the Quasicontinuum Method to Irregular Structures Author(s): *Lars A.A. Beex, University of Luxembourg; Pierre Kerfriden, Cardiff University. The quasicontinuum (QC) method was first developed to reduce computational costs of atomistic computations [1]. It allows one to preserve the full atomistic description in small regions of interest and to coarse-grain the atomistic model elsewhere. As such, it combines aspects of reduced order modelling approaches and concurrent multiscale approaches. For two decades, the QC method was only applied to conservative atomistic models. Not long ago however, it was extended towards discrete dissipative models that can for instance be used to model fibre networks, foams and printed structures [2]. The extension to beam models also enlarged its application range [3]. An important drawback of the QC method is that it can only be applied to lattice models (in which each node/atom is connected to its neighbouring nodes/atoms in the same manner). Work was performed to extend the method towards irregular discrete models and resulted in QC approaches that show similarities to FE2 approaches. Consequently, those extensions come with similar advantages and disadvantages as computational homogenisation approaches. In this presentation we will show a different extension of the QC method towards irregular discrete models. The extension is e ffectively a numerical generalisation of the QC method and has the advantage that the resulting framework remains fully nonlocal. Consequently, the approach can relatively straightforwardly be employed for higher-order interpolation schemes at the macroscale. This will be demonstrated by higher-order examples in which irregular, planar beam models with damage are considered. We consider that the resulting approach has two main advantages. First, no macro-to-micro and micro-to-macro relations need to be de fined. This means amongst others that no boundary conditions at the microstructure need to be applied. This is an advantage because for higher-order interpolation fi elds at the macroscale, these boundary conditions are not trivial. The second advantage is that the separation-of-scales principle does not apply to the method. Hence, the user does not have to worry if scale separation is present. References [1] E.B. Tadmor, R. Phillips, M. Ortiz, "Mixed atomistics and continuum models of deformation in solids", Langmuir, v. 12, p. 4529-4534, 1996. [2] L.A.A. Beex, R.H.J. Peerlings, M.G.D. Geers, "A multiscale quasicontinuum method for dissipative lattice models and discrete networks", Journal of the Mechanics and Physics of Solids, v. 64, p. 154-169, 2014. [3] L.A.A. Beex, P. Kerfriden, T. Rabczuk, S.P.A. Bordas, "Quasicontinuum-based multiscale approaches for plate-like beam lattices experiencing in-plane and out-of-plane deformation", Computer Methods in Applied Mechanics and Engineering, v. 279, p. 348-378, 2014. 14th U.S. National Congress on Computational Mechanics July 17-20, 2017, Montreal, Quebec, Canada Title: An Equation-Free Multiscale Method: Extension of the Quasicontinuum Method to Irregular Structures Author(s): *Lars A.A. Beex, University of Luxembourg; Pierre Kerfriden, Cardiff University. The quasicontinuum (QC) method was first developed to reduce computational costs of atomistic computations [1]. It allows one to preserve the full atomistic description in small regions of interest and to coarse-grain the atomistic model elsewhere. As such, it combines aspects of reduced order modelling approaches and concurrent multiscale approaches. For two decades, the QC method was only applied to conservative atomistic models. Not long ago however, it was extended towards discrete dissipative models that can for instance be used to model fibre networks,