Shadi Alawneh

Hyper-Real-Time Ice Simulation and Modeling Using GPGPU

This paper describes the design of an efficient parallel implementation of an ice simulator that simulates the behaviour of a ship operating in pack ice. The main idea of the method is to treat ice as a set of discrete objects with very simple properties, and to model the system mechanics mainly as a set of discrete contact and failure events. In this way it becomes possible to parallelize the problem, so that a very large number of ice floes can be modeled. This approach is called the Ice Event Mechanics Modeling (IEMM) method which builds a system solution from a large set of discrete events occurring between a large set of discrete objects. The simulator is developed using the NVIDIA Compute Unified Device Architecture (CUDA). This paper also describes the execution of experiments to evaluate the performance of the simulator and to validate the numerical modeling of ship operations in pack ice. Our results show speed up of 11 times, reducing simulation time for a large ice field (9,801 floes) from over 2 hours to about 12 minutes.

Specification-based test oracles with JUnit

Software testing is an important step to help ensure that the software is behaving correctly. An important component of the test process is a test oracle, which determines if the software behavior is correct or not. In this paper, we present tools that enhance an integrated development environment to give the user the ability to write the formal specifications in a readable manner and generate test oracles automatically. The generated test oracles integrate smoothly with test frameworks (e.g., JUnit) and hence they can be directly used to test the behavior of the program. This approach for testing has the advantage that the quality of testing can be high and very efficient.

GPU-Event-Mechanics Evaluation of Ice Impact Load Statistics

The paper explores the use of a GPU-Event-Mechanics (GEM) simulation to assess local ice loads on a vessel operating in pack ice. The methodology uses an event mechanics concept implemented using massively parallel programming on a GPU enabled workstation. The simulation domain contains hundreds of discrete and interacting ice floes. A simple vessel is modeled as it navigates through the domain. Each ship-ice collision is modeled, as is every ice-ice contact. Each ship-ice collision event is logged, along with all relevant ice and ship data. Thousands of collisions are logged as the vessel transits many tens of kilometers of ice pack. The GEM methodology allows the simulations to be performed much faster than real time. The resulting impact load statistics are qualitatively evaluated and compared to published field data. The analysis provides insight into the nature of loads in pack ice. The work is part of a large research project at Memorial University called STePS2 (Sustainable Technology for Polar Ships and Structures). Introduction Ice class vessels are unique in a number of ways in comparison to non-ice class vessels. Hull strength, power, hull form and winterization aspects are all issues that raise special challenges in the design of ice class ships. This paper focuses on matters of local ice loads which pertain to hull strength in ice class vessels. More specifically, the paper examines the parametric causes of local ice loads and statistics that result as a ship transits through open pack ice. The issue of pack ice transit is of interest to those wishing to operate safely in such conditions. One key question is that of safe operational speeds. Consider the special case of open pack ice, where floes are relatively small, numerous and resting in calm water. A vessel moving through such an ice cover would experience a series of discrete collisions. As long as a vessel moved very slowly, the loads would be very low. In such a case the vessel could make safe and steady progress, even if it had a relatively low ice class. However, if the vessel attempted to operate more aggressively, impact speeds would increase and a higher ice class would be needed for safe operations. The investigation below provides some insight into the factors that influence the loads in this situation. These factors include hull form, speed, floe size and concentration, ice thickness, strength and edge shape. Most prior studies have tended to focus on ice thickness and strength as the primary determinants of load. This study shows that ice edge shape and mass, along with hull form and locations are also strong determinants of loads, and especially the load statistics. The simulations provide some interesting data, especially when compared to field trials data. A related focus for the study is to explore the use of the GPU-Event-Mechanics (GEM) simulation approach. The GEM approach represents the integration of a number of concepts. The physical space is described as a set of bodies. The movement (kinematics) of the bodies is tracked using simple equations of motion. Time is divided into relatively long ‘moments’, during which events occur. All variables in the simulation; forces, movements, fractures and other changes, are considered to be aspects of events. Some events are momentary, while others are continuing. Some events involve a single body and are termed solo events. Motion, for example, is treated as a solo event. Some events are two-body events. Impact is an example of a two-body event. The GEM approach lends itself to parallel implementation, which in this case is accomplished in a GPU environment. A GPU (Graphics Processing Unit) is a common element found in modern computer graphics cards. The GPU is primarily intended for making rapid calculations associated with the display. However, special software can access the GPU and enhance the computing power available to the user. See (Daley et.al. 2012) for further discussion of GPUs. The event models are the analytical solutions of specific scenarios. As a result, the events do not require solution (in the numerical sense) during the GEM simulation. The

Ice Simulation Using GPGPU

Simulation of the behaviour of a ship operating in pack ice is a computationally intensive process to which General Purpose Computing on Graphical Processing Units (GPGPU) can be applied. In this paper we present an efficient parallel implementation of such a simulator developed using the NVIDIA Compute Unified Device Architecture (CUDA). We have conducted an experiment to measure the relative performance of the parallel and serial versions of the simulator by running both versions on several different ice fields for several iterations to compare the performance. Our results show speed up of up to 77 times, reducing simulation time for a small ice field from over 88 minutes to about 68 seconds.

Implementation and Performance of a GPU-Based Monte-Carlo Framework for Determining Design Ice Load

Modern Graphics Processing Units (GPUs) with massive number of threads and many-core architecture support both graphics and general purpose computing. NVIDIAs compute unified device architecture (CUDA) takes advantage of parallel computing and utilizes the tremendous power of GPUs. The present study demonstrates a high performance computing (HPC) framework for a Monte-Carlo simulation to determine design sea ice loads which is implemented in both GPU and CPU. Results show a speedup of up to 130 times for the 4 Tesla K80 GPUs over an optimized CPU OpenMP implementation and speedup of up to 8 times for the 4 Tesla K80 over a single Tesla K80 GPU implementation. The elapsed time of the different implementations has been reduced from about 2.5 hours to 0.7 seconds.

Modelling Iceberg-Topsides Impacts Using High Resolution Iceberg Profiles

Platforms operating in arctic and subarctic regions such as the Grand Banks, Labrador Sea, Barents Sea and offshore Greenland are exposed to the risk of iceberg impacts. These structures must be designed to withstand the impact from an iceberg or be designed to disconnect and move offsite to avoid the impact. Offshore Newfoundland, gravity based structures (GBS) such as the Hibernia and Hebron platforms are designed to withstand an impact from an iceberg. However, current accepted practice is not to design the topsides for impact, but to reduce impact risk to an acceptable level by varying the facility geometry (i.e., topsides elevation or footprint). An analytical model was developed to estimate the frequency of icebergs impacting the topsides using three dimensional (3D) models of the platform and the icebergs. Random shapes and sizes are simulated for each iceberg and 3D shapes are generated using a database of measured 2D iceberg profiles. The iceberg shapes are placed randomly in close proximity to the structure and are set to drift towards the structure in a straight line. The initial point of contact between the iceberg and the structure is determined. Crushing of the iceberg against the platform caisson is considered. The process is repeated a large number of times and the total number of contacts with the topsides are determined. In 2012, Hibernia Management and Development Company Ltd. (HMDC) sponsored a field program in which high resolution iceberg profile data were collected. The high resolution iceberg profiles contain detailed 3D information of the above water and below water shape of the iceberg. This paper describes updates to the existing-topsides impact model to take full advantage of the detailed 3D iceberg profiles. These updates include new iceberg shape databases for simulation, and the addition of a detailed iceberg management model and a graphical user interface (GUI) to improve the functionality of the software. Introduction Icebergs can pose a significant risk to oil and gas exploration, development and production facilities operating on the Grand Banks, off Canada’s east coast. The Terra Nova and White Rose floating production storage and offloading (FPSO) vessels are designed to disconnect and move off location to avoid impacts for icebergs which cannot be managed. The Hibernia GBS was designed with an outer ice wall capable of resisting impacts from large icebergs (Hoff et al. 1994; Huynh, Clark, and Luther 1997). The Hebron GBS is also designed to withstand impacts from large icebergs (Widianto et al. 2013). With both GBS platforms, the caisson can withstand large impact forces. However, it is not practical or feasible to design the various components of the topsides structure (e.g. walkways, lifeboat stations, generators, etc.) to withstand such large impacts forces. Instead, the topsides layout and elevation above the sea surface are designed such that the risk of iceberg impact is minimized. The Iceberg-Topsides Impact Model was developed to provide guidance to designers. This software tool can be used during the early concept selection stage of a project to optimize the basic GBS-topsides configuration such that the risk of an iceberg impact with the topsides is minimized. The tool can also be used later during detailed engineering design to verify that a particular design meets ISO 19906: 2010 guidelines. During the summer of 2012, HMDC sponsered an extensive field program to collect high resolution iceberg profiles (Younan et al. 2016). Multibeam sonar was used to collect below water iceberg shape information and a photogrammetry system was used to collect above water information. The data were combined resulting in complete 3D iceberg profiles.

Accelerating Numerical Ice Engineering Tools Using GPGPU

C-CORE is engaged in understanding the iceberg and sea ice design loads needs of the energy sector. As the energy industry ventures into oceans with greater ice cover and more icebergs, there is a significant need for efficient engineering tools to plan and manage operations in exploration, production, and safety. Industry requires a range of scenarios for their risk assessments, where existing simulations can be computationally and time intensive. C-CORE has recently started using the benefit of the General Purpose Computing on Graphical Processing Units (GPGPU) approach. This approach has shown significant speed up of several numerical ice engineering applications related to icebergs and sea ice. The investigated model types are Monte-Carlo type approaches for probabilistic design method, and quadratic discriminant. GPU computing with Compute Unified Device Architecture (CUDA) is a new approach to solve complex problems and transform the GPU into a massively parallel processor. The present study applies the GPGPU technology to a Monte-Carlo simulation, used for a sea ice load application. The objective of this study is to measure the performance of the GPU using CUDA, and compare against the serial Central Processing Unit (CPU) using C++ and MATLAB implementations. Results show a speedup of up to 2,600 times of the GPGPU implementation compared to the MATLAB implementation, reducing the elapsed time from about 1.5 hour to about 2 seconds. This strongly indicates that the GPGPU approach can help the industry to significantly reduce the time required for the simulations.

2D Triangulation of Polygons on CUDA

General Purpose computing on Graphics Processor Units (GPGPU) brings massively parallel computing (hundreds of compute cores) to the desktop at a reasonable cost, but requires that algorithms be carefully designed to take advantage of this power. The present work explores the possibilities of CUDA (NVIDIA Compute Unified Device Architecture) using GPGPU approach for 2D Triangulation of Polygons. We have conducted an experiment to measure the performance of the GPU with respect to the CPU. The experiment consists of implementing a serial and a parallel algorithm to triangulate 2D polygons and executing both algorithms on several different sets of polygons to compare the performance. We also present an application that uses the polygon triangulation.

USING TEST ORACLES AND FORMAL SPECIFICATIONS WITH TEST-DRIVEN DEVELOPMENT

This paper illustrates how Test Oracles and Formal Specifications, with appropriate tool support, can be used with Test-Driven Development (TDD). In TDD, the test code is a formal documentation of the required behavior of the component or system that is being developed, but this documentation is necessarily incomplete and often over-specific. We describe an alternative approach to TDD that is to develop the specification of the required behavior in a formal notation as a part of the TDD process and to generate test oracles from that specification. We present the results of using this approach to develop programs used in a project at the Faculty of Engineering and Applied Science at Memorial University.