Components
The final design for this project contained two major components – the cradle and the frame. The cradle was comprised completely of steel. Two U-channels of 144.78 cm in length formed
the height of the cradle. Steel tubes connected the two U-channels. The server cabinets contain a lip near the bottom of the frame; these lips are reinforced for earthquakes. Therefore, the bottom of the cradle was designed to latch onto this lip in order to attach the cabinet to the frame
for tilting and transport. Because two lip-grabbing components were to be used at the base of the cradle, this design was well within yield strength parameters. In addition to this lip-grabbing mechanism, two straps (each with a width of 10.16 cm) wrap around the cabinet to provide extra
support during tilting. Each strap has a 2267.962 kg (5000lb) capacity. The other major component of this design was the frame. Steel tubes were welded together to create the exterior portion of the frame. Tubing was used in order to create outriggers that extended from the tubes.
These outriggers can be viewed in Figure 3. Because the cabinet was intended to be placed in the front of the device, an excess amount of weight would cause the device to tip over. Therefore, outriggers were designed in order to avoid a moment in the front of device.
Design Process
The first step in the design process was to research current designs. Initial research led to a design that consisted of modifying an existing high-capacity dolly. However, this idea was thrown out after analysis showed this would be difficult. Research then lead to the idea
of modifying upenders. However, these were far too heavy for the 362.874 kg (800 lb) constraint. While researching updenders,one design demonstrated a tilt table that was mounted onto a rolling frame. After deciding the upenders on the market did not allow for much
trans portability, a build-from-scratch approach was taken as an inspiration for an overall design.The team decided to split the design into two major components – the frame and the cradle. The major breakthrough for the design came after discussing the possibility of a moment
being caused at the front of the device from the weight of the cabinet. The solution to this moment problem was the use of outriggers. It was then decided to use steel tubing to allow for inner tubes to slide out for more stability. The placement of the hydraulics was one of the most
discussed features during the design process. Research of existing hydraulic-powered tilters showed two main set-ups for hydraulics – bottom-driven and top-driven. A bottom-driven hydraulic system simply means that the hydraulics push and pull at the base of the tilting object.
A top-driven model places the hydraulic at an angle to induce tilting. Figures 18 and 19 in the appendix demonstrate the analysis for both the bottom-driven and top-driven models. Using the analysis, and through a number of tests using the wood model, a bottom-driven approach was
decided upon. The bottom-driven approach in the wood model demonstrated smoother tilting than the top-driven approach (Figure 3). Many designs were discussed for the cradle. Initially, forklift forks were going to be used for the cradle. However, it was eventually discovered that
these would be too heavy. A lip-grabbing cradle as shown in Figure 5 was then decided upon. The two major components would be connected using a four-bearing shaft design as seen in Figure 6.
Table 3: Final Design Features. This table shows the full range of motion of the tilting mechanism.
Figure 6. Bottom Driven Design
Bottom Driven Design
Figure 7. Full 90 Degree Tilt
Full 90 Degree Tilt
The objectives of this project led to high demands during the fabrication process. Many components were bought “off-the-shelf” in order to make reproduction much easier. Most of the components described in Table 4 were purchased rather than fabricated. Rather than fabricating
one section of the design at a time, fabrication occurred as parts came in. Because parts were ordered from a number of distributors, parts arrived in random fashion. The cradle can be seen in Figure 3. Two large, steel u-channels make up the height of the cradle. Steel cross sections
provide support. The parts of the cradle were connected via welding. The majority of the difficulties came during the fabrication of the frame (Figure 6). Steel tubes were welded together for exterior of the frame. Plates were placed between the casters and the bearings. These plates had
to be chamfered and tapped. The bearings with plates were welded onto the frame. The intricacy of the frame fabrication is apparent around the shaft area (Figure 6). The relation of the hydraulics, the bearings, and their connection to the cradle was one of the most challenging components
of fabrication.
Figure 8. Final Design After Fabrication.
After completing the fabrication, testing commenced immediately. The testing plan is listed as follows:
Rotate Cradle from 90° to 0° (Vertical to Horizontal) and 0° back to 90° with:
1. Empty Server Cabinet (136.1 kg (300 lbs)
2. Cabinet + 4 Servers (453.6 kg = 1000 lbs)
3. Cabinet with all Servers supplied by Teradata (589.7 kg = 1300 lbs)
Figure 9. Design in Action. This shows the full assembly and tilt motion with the cabinet attached.
Figure 10. Testing Results. This table demonstrates the speed of tilting at three different loads.
The results demonstrate a drastic improvement to how Teradata originally had to tilt their cabinets through doorways. Rather than having to empty the cabinet and then placing it at a horizontal position on dollies to transfer it through doorways, this design allows for the full
tilting motion with servers installed in a matter of minutes.