IoT Workload Emulation for Data Centers.

The Internet of Things (IoT) domain has been one of the fastest growing areas in the computer industry for the last few years. Consequently, IoT applications are becoming the dominant work load for many data centers. This has implications for the designers of data centers, as they need to meet their customers' requirements. Since it is not easy to use real applications for the design and test of data center setups, a tool is required to emulate real applications but is easy to configure, scale and deploy in a data center. This paper will introduce a simple but generic way to model the work load of typical IoT applications, in order to have a realistic and reproducible way to emulate IT loads for data centers. IoT application designers are in the process of harmonizing their approaches on how architectures should look, which building blocks are needed, and how they should interwork. While all IoT subdomains are diverse when it comes to the details, the architectural blueprints are becoming more and more aligned. These blueprints are called reference architectures and incorporate similar patterns for the underlying application primitives. This paper will introduce an approach to decompose IoT applications into such application primitives, and use them to emulate a workload as it would be created by the modeled application. The paper concludes with an example application of the IoT Workload Emulation in the BodenTypeDC experiment, where new cooling approaches for data centers have been tested under realistic work load conditions.

Comparison of a fixed-grid and arbitrary Lagrangian-Eulerian methods on modelling fluid-structure interaction of the aortic valve.

This study was aimed at assessing the robustness of a fixed-grid fluid-structure interaction method (Multi-Material Arbitrary Lagrangian-Eulerian) to modelling the two-dimensional native aortic valve dynamics and comparing it to the Arbitrary Lagrangian-Eulerian method. For the fixed-grid method, the explicit finite element solver LS-DYNA was utilized, where two independent meshes for the fluid and structure were generated and the penalty method was used to handle the coupling between the fluid and structure domains. For the Arbitrary Lagrangian-Eulerian method, the implicit finite element solver ADINA was used where two separate conforming meshes were used for the valve structure and the fluid domains. The comparison demonstrated that both fluid-structure interaction methods predicted accurately the valve dynamics, fluid flow, and stress distribution, implying that fixed-grid methods can be used in situations where the Arbitrary Lagrangian-Eulerian method fails.

Multiphysics simulation of the effect of leaflet thickness inhomogeneity and material anisotropy on the stress-strain distribution on the aortic valve.

This study developed a realistic 3D FSI computational model of the aortic valve using the fixed-grid method, which was eventually employed to investigate the effect of the leaflet thickness inhomogeneity and leaflet mechanical nonlinearity and anisotropy on the simulation results. The leaflet anisotropy and thickness inhomogeneity were found to significantly affect the valve stress-strain distribution. However, their effect on valve dynamics and fluid flow through the valve were minor. Comparison of the simulation results against in-vivo and in-vitro data indicated good agreement between the computational models and experimental data. The study highlighted the importance of simulating multi-physics phenomena (such as fluid flow and structural deformation), regional leaflet thickness inhomogeneity and anisotropic nonlinear mechanical properties, to accurately predict the stress-strain distribution on the natural aortic valve.

The effect of cerebrospinal fluid thickness on traumatic spinal cord deformation.

A spinal cord injury may lead to loss of motor and sensory function and even death. The biomechanics of the injury process have been found to be important to the neurological damage pattern, and some studies have found a protective effect of the cerebrospinal fluid (CSF). However, the effect of the CSF thickness on the cord deformation and, hence, the resulting injury has not been previously investigated. In this study, the effects of natural variability (in bovine) as well as the difference between bovine and human spinal canal dimensions on spinal cord deformation were studied using a previously validated computational model. Owing to the pronounced effect that the CSF thickness was found to have on the biomechanics of the cord deformation, it can be concluded that results from animal models may be affected by the disparities in the CSF layer thickness as well as by any difference in the biological responses they may have compared with those of humans.

The importance of fluid-structure interaction in spinal trauma models.

While recent studies have demonstrated the importance of the initial mechanical insult in the severity of spinal cord injury, there is a lack of information on the detailed cord-column interaction during such events. In vitro models have demonstrated the protective properties of the cerebrospinal fluid, but visualization of the impact is difficult. In this study a computational model was developed in order to clarify the role of the cerebrospinal fluid and provide a more detailed picture of the cord-column interaction. The study was validated against a parallel in vitro study on bovine tissue. Previous assumptions about complete subdural collapse before any cord deformation were found to be incorrect. Both the presence of the dura mater and the cerebrospinal fluid led to a reduction in the longitudinal strains within the cord. The division of the spinal cord into white and grey matter perturbed the bone fragment trajectory only marginally. In conclusion, the cerebrospinal fluid had a significant effect on the deformation pattern of the cord during impact and should be included in future models. The type of material models used for the spinal cord and the dura mater were found to be important to the stress and strain values within the components, but less important to the fragment trajectory.

Poisson's ratio and strain rate dependency of the constitutive behavior of spinal dura mater.

Knowledge of the mechanical behavior of spinal dura mater is important for a number of applications including the experimental and computational modeling of physiological phenomena and spinal cord trauma. However, mechanical characterization of dura mater is relatively sparse and is further compounded by the use of the tangent modulus as the sole measure of stiffness. This study aims to provide a more complete description of the mechanical properties of spinal dura mater, including the effect of strain rate. Bovine dura mater was tested under uniaxial tension in both the longitudinal and the circumferential directions at three different strain rates; 0.01, 0.1, and 1.0 s(-1). An Ogden model was fitted to the resulting stress-stretch data. The morphology of the dura mater was assessed using Sirius red and H&E staining. No significant effect of the strain rate was found for the Ogden model parameters. Longitudinal specimens were significantly stronger and more deformable than circumferential samples, probably due to the structural arrangement of the collagen fibers. At low strains, however, the circumferential specimens were stiffer than the longitudinal ones. The findings of this study will allow more complete representations of the spinal dura mater to be developed.

The effect of bone fragment size and cerebrospinal fluid on spinal cord deformation during trauma: an ex vivo study.

OBJECT: The purpose of the study was to assess the effect of CSF and the size of the impacting bone fragment area on spinal cord deformation during trauma. METHODS: A transverse impact rig was used to produce repeated impacts on bovine and surrogate cord models. Tests were recorded with high-speed video and performed on specimens with and without CSF and/or dura mater and with 3 different impactor areas. RESULTS: The CSF layer was found to reduce the maximum cord deformation significantly. A 50% reduction in impact area significantly increased the maximum cord deformation by 20-30%. The surrogate model showed similar trends to the bovine model but with lower absolute deformation values. CONCLUSIONS: Cerebrospinal fluid protects the cord during impact by reducing its deformation. A smaller bone fragment impact area increases the deformation of the cord, in agreement with clinical results, where a higher impact energy-possibly giving rise to smaller fragments-results in a worse neurological deficit.