(209a) Mining Big Data from Product & Process Simulations to Improve Designs

Advancements in computer modeling & simulation, also known as virtual prototyping,

have already dramatically reduced the need for exhaustive physical prototyping and testing

of complex subsea subsystem, while concomitantly increasing the understanding of the

proper and safe operating ranges of such equipment.  While these advancements have provided

a significant improvement in speed-to-market and an associated reduction in testing costs,

there remains a huge untapped opportunity to also reap major product design improvements by 

better leveraging the potential of virtual design exploration.

            In this paper, we will demonstrate how it is now possible to leverage three available technologies

to mine massive simulation data and discover better subsurface and subsea designs, faster than ever before.  The 

combination of robust computer-aided engineering simulation models, coupled with ubiquitous

high performance computing resources, and innovative direct design search strategies enable

radical improvement in the quest for designs that perform better, cost less, and operate more safely

for a longer period of time.  We will demonstrate the effectiveness of this big data mining strategy

in the design of subsea heat exchangers, off-shore risers, and oil drilling equipment.