This is part six of a 10-part series chronicling the R&D of a wave energy converter. Read parts one, two, three, four, and five.

With the big yellow foam buoy already built, our team of engineering students now focused on building the long slender spar assembly. The spar is attached to the underside of the buoy and is the frame that keeps the hydraulic cylinder and mechanical spring assembly aligned with each other.

It’s also the critical feature that allows for the relative motion between the buoy and heave plate to push and pull on the hydraulic cylinder as waves move past the wave energy converter (WEC). This cyclic pushing-pulling motion pumps the hydraulic fluid that turns the hydraulic motor and spins the generator to produce electricity. In keeping with the objective to make the WEC easy to move and deploy, the spar was broken down into three sections so that the long heavy structure could be dissembled for transport.

The spar was designed to be as short as possible; this was done to minimize the overall weight and reduce the cost of materials. To accomplish this, each section was designed to be a different length in order to accommodate the various equipment and parts that had to fit in each module. The hydraulic cylinder and compression bushing were built into the 79″-long section; the mechanical springs, polymer alignment collars, and steel extension shaft were built into the 90″ section; and the high-density polymer bushing and steel cable were installed in the 60.5″ section. This meant that each spar section could be assembled independently from one another and transported as pre-assembled units. The three sections simply bolted together to create one long structure using 16 Grade 8 bolts.

Final layout of spar assembly with part dimensions and names.

Final layout of spar assembly with part dimensions and names.

The spar was built out of square steel tubing that was 2″ wide and 1/4″ thick. While this may seem over-engineered, this was done to prevent the spar from failing in the case of a freak storm with gigantic waves. In an ideal engineering problem, it’s possible to predict or even measure the forces that will be applied to your design. This information helps the engineer select the specific type of material needed to prevent the part from failing and is one of many factors (including cost, weight, machinability) that shapes the final product. However, as we were learning the hard way, real-world applications are not always so simple and therefore the team was apprehensive about trying to predict the maximum forces that would be exerted on the WEC. Luckily, we had learned about a special design criteria used for building mechanical systems that had to survive the extreme marine environment, the Francois Factor. Definition: Start by designing the system to be as strong as you think it will need to be in order to survive in the ocean, then make it FIVE times as strong. (Affectionately named after Mr. Francois Cazenave, who offered this advice during our MBARI trip.)

WEC_Normal opperation

We needed to define a worst-case scenario and approximate the maximum upward force on the spar before applying the Francois Factor. That was relatively simple. We assumed a massive wave would move past the WEC and cause the hydraulic cylinder to fully elongate, at which point the buoy would continue to be pulled upwards while the heave plate would be pulled downwards due to the hydrodynamic drag. This would place the spar in maximum tension. To approximate this tensile force, we took the total upward buoyant force of the buoy, 3800 pounds, and assumed that this would be the maximum pull exerted on spar.

This assumption was valid as long the heave plate remained fixed near the bottom of the ocean, and was a simplification we used throughout the analysis. We also neglected the weight of the spar, as it would only act to decrease the maximum buoyant force of the buoy. In reality, to generate a downward hydrodynamic drag force, the heave plate must be moving upwards with some velocity, but we made an underlying assumption to simplify the analysis and assumed that the heave plate acted as an ideal “anchor.” In this way we could assume that all the upward buoyant forces generated by the buoy would be transferred to eight bolts connecting the spar sections together.

Diagram of assumed worst case senario for WEC

Diagram of assumed worst-case senario for the WEC.

With the worst-case scenario worked out, we then applied the Francois Factor to calculate a total tensile load equal to 19,000 pounds. Knowing that we had initially planned to use eight bolts to connect each of the sections together, it was then possible to calculate the stress in each bolt and see if that value was less than the 120,000psi proof load of the Grade 8 bolts. The results from all of this indicated that the spar would be 10 times as strong as it needed to be (assuming that our worst-case scenario was valid) and since the cost of the extra-strength bolts was not that much more compared to the other options we stuck with the original plan and built a really strong spar. This same thought process and procedure was used to determine the size and thickness of the square steel tubing used to build the spar sections.

Cleaning the Steel

The steel stock arrived in 20-foot-long sections and was ordered from a local vender in Sacramento, Calif., who delivered it to the school shop on a flat-bed truck. Once all 180 linear feet were unloaded from the truck, the team set to work cleaning the surface of the tubing. During the manufacturing and processing at the steel mill, the surface of the steel tubing is covered in a black oil to protect the steel from moisture and prevent rusting. While the oil helps to protect the steel while it’s in storage, the same oil can cause issues during fabrication and therefore must be removed before welding or painting. We sprayed WD-40 on the steel, which helped to break down the black oil and make it easier to wipe off with rags.

This slideshow requires JavaScript.

Cutting the steel to size

Once the surface had been somewhat cleaned, it was time to measure and mark the 20-foot-long pieces of steel tubing and cut them to size with the massive vertical bandsaw. We were lucky to have access to such a large and powerful machine. With the steel tubing being 1/4″ thick, this step took a long time and someone always had to be watching the machine to ensure nothing went wrong.

This slideshow requires JavaScript.

Grinding the steel and adding chamfers

After the steel had been cut, the team use angle grinders to grind away any remaining oils and contaminates around the edges of the tubing. Before welding the pieces together, the area that was to be welded had to be extremely clean. This grinding process was a quick and efficient way to prepare all the surfaces. Chamfers were also added to the ends. This way, when two pieces of tubing were placed together at a 90° angle there would be a small valley-created V-groove that would then be filled with welding filler material. This V-groove technique creates a stronger weld and allows for a larger amount of material to be added to the joint, ensuring deep penetration and optimal welding conditions in corner regions.

This slideshow requires JavaScript.

Drilling Holes

The last step before welding was to drill two sets of 5/8″ holes through each of the 24 crossbars. Team members Alex Beckerman and Tom Rumble set up an assembly line using two of the shop’s Bridgeport vertical knee mills, and drilled all the holes over the course of a few afternoons. With the relatively thick steel tubing, the drilling process was slow. The location of the holes was also critical and this meant that Alex and Tom had to be very careful when placing the parts in the vice.

This slideshow requires JavaScript.

Welding the Frames

The ends of the square steel tubing were cleaned one more time with a clean rag before we started welding; if there was any oil or dirt at the surface, it could cause voids and defects that would reduce the strength of the weld. To start, we created a box with four pieces of tubing and carefully clamped everything down to the table. After measuring and adjusting the pieces to ensure that they were as square as possible, we then added tack welds at the corners using a MIG welding machine.

This slideshow requires JavaScript.

With two of these boxes built, we then flipped them up on their side and added four more 8″-long cross sections and additional clamps before tack welding the remaining connection points. After measuring everything one final time to make sure the frame was as square as possible, we proceeded to weld everything together. If any of the three steel spar sections were crooked or skewed, this would cause uneven loading during operation and potentially lead to premature failure. Making the sections square and uniform was extremely important.

Due to the thickness of the steel and the lengths of each weld, we had to be careful to evenly distribute the heat generated during the welding process so as to avoid warping of the frames. If we had welded all of the seams in one corner all at once, then that one portion of the steel frame would have been much hotter than the other three corners and cause the frame to warp due to thermal stresses. We had to work in teams with one person welding 2″-long seams while two other people would flip the heavy frames so the welder could get access to a different corner that had already had time to cool down.

Square plate welded to the end of each frame.

Square plate welded to the end of each frame.

We also welded 1/4″-thick steel plates flush with the ends of each frame. These plates had been cut out of a larger 4’x4′ sheet of steel using the shop’s CNC plasma cutter and already had the eight bolt holes cut out around the perimeter. There was also a large hole in the center of the plates for the hydraulic cylinder, springs, and steel cable to pass through. Once the end plates had been added, the frames were allowed to fully cool before being lifted off the table.  Finally, we used a large industrial scale to measure the weight of each section and verify they where under the allotted weight limit for the design.

Assembling the Spar

After gently flipping the buoy on its side, we quickly realized that we needed to clear more space in the student shop before we could test-mount the 19′-long spar assembly. Having reconfigured the workspace around us, we tried to attach the 79″-long frame to the bottom of the buoy and realized that we had a problem. To attach the spar to the buoy and create a rigid connection point, we had planned to bolt the entire setup to the bottom of the buoy using eight 1/2″-13 threaded rods. The spar would be butted up against the bottom steel plate of the buoy and bolted in place.

This slideshow requires JavaScript.

However, one of the eight threaded rods was sticking out at an angle due to a defect with the buoy. Because of this slight angle, it was not possible to align all eight of the threaded rods with the holes drilled into the top of the spar, thus making it impossible to mount the spar flush with the bottom of the buoy. After trying to enlarge the hole in the steel tubing and bend the rod into place, we ran the strength calculations one more time (the same calculations used for the Grade-8 bolts mentioned previously). This verified that seven rods would still be able to support the load of our assumed “worst-case scenario” so we decided to remove the rod altogether.

This slideshow requires JavaScript.

Having removed the bent rod from the buoy, we then used a set of ratcheting tie straps to bring the spar flush against the bottom of the buoy, at which point it was possible to add the washers and hardware. The other two remaining sections of the spar went on without any issue and just bolted together like we had planned. The WEC was now nearly 25′ long from the top of the buoy to the bottom of the spar and was catching the attention of everyone who walked into the shop. By this time the team was feeling confident. We had accomplished quite a lot in the first five weeks of building. But now we ran into a new kind problem that I personally had never had to deal with before: what do you do when you need to move a 1,200-pound object to the other side of the shop?