Introduction

Our Group was tasked to find a solution for the reduction of Dynamic Amplification Factor (DAF) during offshore lifting that required us to understand more of this industry. During offshore lifting, an object is rigged onboard and carried down to the seafloor by means of a vessel crane. When it passes through the splash-zone, it experiences hydrodynamic forces that cause the object to undergo heave and pitch motion. This causes the lifting slings to undergo alternating motions of loosening and tensioning. Thus a dynamic amplification factor (DAF) is applied to the physical system so that the possibility of sling snap does not occur. This value of DAF can be obtained from the DNV code (DNV-RP-H103). DAF can be calculated by dividing the sum of the static and hydrodynamic forces. The hydrodynamic force is made up of 1) Drag Force acting on the object, 2) Varying Buoyancy Force due to wave surface elevation, 3) Mass Force due to combination of inertia forces and 4) Slamming Force which acts vertically on the bottom of the object.

The host company had commented that the calculation of the DAF based on the DNV code is too conservative. Therefore, our objective for this project is to provide a better and more accurate estimation of the hydrodynamic forces, proving that the DNV formulations are too conservative. By running a simulation of this lifting operation, we are able to obtain necessary values of the different forces to proceed with our calculation of DAF, and thus able to compare it with calculation based on the DNV code. Our group has adopted FLUENT, a computational fluid dynamic software that is hosted by HPC, as the platform for us to perform our analysis.

Experiences with HPC resources

The HPC team had provided us with ample resources to achieve our objective. We can sum this up into three components – online FLUENT guide, Secure Shell (SSH) and parallel processing.

Having no prior knowledge to any computational fluid dynamic (CFD) software, the online FLUENT guide had provided us with a venue to begin our studies on CFD which we had to be proficient in, in order to yield useful results for our project. In addition, there were also many online tutorials that we could download for practice. These online tutorials provided step-by-step instructions that were easy to follow and the tutorials came in a great variety ranging from pipe flow to enthalpy change due to fuel combustion. Thus, the group was able to select tutorials that were similar to our case and practice on them before starting on our project case.

Secondly, with the availability of SSH, the group was able to store our saved projects on the Active Directory. We were then able to access and modify our saved files with the download of the SSH client on our personal computers. Thus, this convenience allows the group for greater flexibility when delegating work and easy accessibility to our saved files anywhere.

Lastly, paralleling processing is another important component for our simulations. Having completed the configuration of our solver, the next step was to summit our job to the Load Sharing Facility (LSF) for processing and this is done through the xterm. Within the xterm, we were able to configure options such as the number of CPUs to be used during processing and define settings to automatically save images of our simulations during regular intervals. Furthermore, we were able to monitor the progress of our simulation easily on the xterm. Thus, with the use of parallel processing, it speeds up the process of the simulation while enabling us to check its progress conveniently without having to keep a computer on to run the simulation.

Benefits

The online FLUENT guide was helpful in our process of figuring out the know-hows of the software. Coupled with the online tutorials provided, we were able to build a foundation for our project. Furthermore, the SSH Shell Client allowed us to have access our saved files in the Active Directory from both in and out of campus, which provided with great convenience when working on the project. It also enabled us to track the progressive results of our simulation; thereby allowing us to have a good gauge of our job. To sum up, the resources that HPC had provided us was a crucial aid in the building of our knowledge of the software while other software such as SSH enabled us to work anywhere with great convenience.

Recommendations

All in all, the resources provided by HPC were extensive and very useful for our project. To further benefit users who are also making use of the HPC resources, we suggest that a more extensive FAQ could be included. Other than that, alternative resources such as guidebooks that are available online or books can also be recommended.

One issue our group have faced with was that although there was plenty of information, ranging from software usage to running of jobs that is available on the HPC website, they were not easy to locate. We often had to search through the website to locate information that we were seeking, and the HPC specialist whom we have consulted often had to point out the link to the information we requested for. Hence, we suggest that there could be a one-stop access page that includes a collation of all the links to the information. It could be a step-by-step guide, where users are guided from the very start of software usage to the very end; from the creation of account to the running of the job.

In addition, the staff could also compile their own guides that are simple to read, understand and be applied. Examples for such guides are coding and troubleshooting guides. The guide should be extensive because online materials were scarce and though we often followed them sequentially, we could not get the desired result. Furthermore, some software books were too complex to understand. Thus, compilation all the possible condition and scenarios categorically for easy referencing will enable students to learn faster, and avoid errors hence troubleshooting, and work can be done faster too.

Finally, a link to the blog could be explicitly shown in SVU VisLab for students to be better aware of this blog. For example, provision of details on the wallpaper of computer screen display in the VisLab so that students can easily get access to them. Furthermore, contacts of HPC specialists may also be made available in VisLab such that the VisLab may become a one stop hub for all these information.

Conclusion

The experience with the HPC resources was a fruitful one. We were able to learn and utilise complicated CFD software such as FLUENT to run our case as well as to yield promising results for our project. Facilitated with many other resources that the HPC team had provided, we were able to manage our simulations in and out of campus with great convenience. As initial users might be overwhelmed by the vast amount of software available at their disposal, we suggest a comprehensive guide that contains step-by-step instructions especially in areas such as job submission. Furthermore, more information can be put up in the SVU VisLab to facilitate users, be it contacting a HPC specialist or for self-help. Lastly, the group will like to thank HPC specialist Dr Wang Jun Hong for his great patience and his invaluable assistance for our project.