The purpose of these watchful guardians became clear as we observed them “rescue” stranded tube-riders, one after another, from pools, eddies, and other trouble spots, giving them the push they needed to move them downstream and through the ride. Though charming, the west park clearly lacked the planning and efficiency that was enjoyed, no doubt by tubers and lifeguards alike, in the rides of the east park. This likely stood out more to me than the average park-goer because of my involvement in a numerical modeling project for a whitewater rafting and kayaking run, while working at Aquaveo.
S2O Design, the design firm for RIVERSPORT Rapids—which is currently being built in Oklahoma City, OK—was building a small-scale physical model of the site and wanted results from a numerical model to better understand how the rapids would work and to ensure the runs provided the best possible experience for the riders. Serving both recreational riders and Olympic hopefuls, the venue includes two runs—a faster, more challenging competition run and a slower, more leisurely recreational run. Although the runs are separate from each other, they share the same upper and lower pools, which created a unique challenge to come up with a one-size-fits-all solution for these areas.
The numerical modeling project that took place over several months was complex, challenging, and exhilarating. Over the course of developing the project, I learned several valuable lessons:
Lesson 1: Before you go big, you gotta go small
The success of a large scale, technically challenging project often relies on science performed on the small screen. A numerical model gives an engineer the ability to see the hydraulic impacts of a proposed design concept before the designs are constructed. It also lets an engineer tweak the design and rerun the analysis without the high costs of brick and mortar changes.
This was the case with the design of the whitewater venue. For all the fast-moving currents and whitewater rapids, the biggest concern was the entrance pool, where the motion of riders relies more on level currents than on gravity. A well designed entrance pool lets water expand to form still water areas while maintaining a slow, steady current towards the launching area. A bad design could either sweep riders away too fast or trap them in eddies causing jams at the entrance. For the whitewater model it was the latter, and a subtle change to the angle of a sidewall was all it took to fix—a small, simple change in a computer model that paid big dividends to the success of the course.
Lesson 2: Don’t assume the most expensive tool is the best, test
The federal government has had a bad track record with their 2D models in the past. But having an open mind and testing the US Bureau of Reclamations Sedimentation and River Hydraulics in Two Dimensions (SRH-2D) served me well. I was able to compare the results of the computer model with measurements obtained in the small-scale physical model to gauge accuracy. I tested two other proprietary models for comparison and found that SRH-2D was the most accurate and had the fastest run-time (and was also cheaper and easier to use). Although proprietary models still have their place and always will, it was evident from this study that government-developed 2D models have found their place too—and it’s not just in academia. It’s important to continuously test a variety tools for best performance. Sometimes you are pleasantly surprised.
Lesson 3: When models are wrong, you should know why
Velocity Contours Model - courtesy of Aquaveo
While testing, I discovered all 3 numerical models were reporting much higher velocities than the physical model, although everything else was testing out well. With limited time and budget, I decided not to invest the time to truly understand why the velocity results were not matching up in order to focus on what I was able to verify. I was asked to give a conference presentation on the project, and made a fair attempt to explain the velocity discrepancy without dwelling on it. But when I got to the Q&A session, my heavily academic audience was less than forgiving. It was an uncomfortable lesson, but I now know that it’s okay if a model is wrong, as long as you know the reason why. And if you don’t know the reason why, it’s better to state that, than to glance over. Over the next couple months, I dug into several journal articles and found the answer I was looking for and by the time I presented the study at a separate conference several months later, dissatisfied raised hands were replaced by approving nods. I learned small discrepancies are not to be overlooked—they are there to teach us big lessons.
"Design for whitewater parks is fraught with uncertainty. The design of turbulent flows has to be tailored to the specific tastes of the users to create waves, hydraulic jumps, eddies, and jets that cater to surfers, kayakers, and rafters of all levels. 2D models provide a great way to visualize actual features and the finer nuances, allowing us to optimize our design before creating the physical model."
Scott Shipley, M.S., P.E., President/Owner, S20 Design and Engineering
The numerical modeling project was successful and provided important feedback that eliminated trouble spots in the physical model, and ultimately contributed to a better overall design for the whitewater venue—confirming flow patterns in the physical model and eliminating eddies in the entrance pool. Not only will the project allow lifeguards to enjoy the excitement from a perched tower instead of standing waist deep in the water, it has opened doors to new possibilities for analyzing complex hydraulics accurately and within budget.
The park is set to open in March 2016 and will feature canoeing, kayaking and rafting for Olympic trainers and the general public. It will also be used for water rescue training. Just as this park taught me a few lessons, I’m sure it will also teach some aspiring athletes and water rescue professionals a few lessons as well. It’s sure to be a treasured feature of Oklahoma City for decades to come.
Watch the small-scale physical model video reproduced here with kind permission from S20 Design and Engineering: