This was the second major project I worked on during my internship at Shawcor Composite Production Systems. The objective was to design a new rack that would allow for easier manual manipulation of composite pipes by R&D department staff. The R&D department frequently need to access pipe segments of samples for testing purposes. The racks normally used in the facility (shown below) do not allow access to individual pipes, only to those on top. It is also very difficult to move pipes by hand on the rack because they are extremely heavy. 

The design of a new kind of rack required:
1) stored pipe to be easily moved by by hand while the pipe remained on the rack

2) the rack to accommodate being moved around the facility by forklift

3) able to have multiple racks stacked on top of each other

To easily move pipes by hand, it required some sort of low friction device, such as a roller. Desired axes of movement are rotation and translation, with translation being more important. The first potential solutions considered were the various rollers used in the facility. Some rollers only allowed for rotational movement, and were therefore unacceptable given the requirements. Other multi-axis rollers in the facility were considered. When I considered the number of rollers needed for the rack, it became apparent that using any of these other rollers would drive the price of the rack up significantly. Below are shown some of the rollers used in the facility. 

Since translational movement was most important, a simple boat roller was considered as a solution. This was a desirable solution because they are very inexpensive. 

Image credit: https://www.princessauto.com/en/detail/12-in-keel-roller/A-p8558132e 

To see if the boat rollers were a good solution, I wanted to check for their ability to keep pipes centered, tendency to deform under load, and positioning of the rollers. This was done by creating a proof of concept jig out of supplies from a local hardware store. Based on this test, the boat roller was deemed acceptable. 

After considering a few options for the structure of the rack, it was initially narrowed to either an E-Z-Rect-style of rack (shown below) modified to use rollers, and a custom welded steel structure. The E-Z-Rect racking was determined to be not easily moved by forklift in the Shawcor facility. So, a custom steel welded structure was the obvious solution. 

Image credit: https://www.commander.ca/product-ezrect-shelving-type-1/

The next steps involved further technical requirements gathering. The most important of these requirements were that the racks needed to be able to fit on the facility's active forklifts. They also could not pose a tipping risk to the forklifts during transportation. The combination of the rack's depth and estimated weight were compared to the lift capacity of the forklifts to determine geometric constraints on the racks. 

The racks were designed to accommodate the size and weight of the various pipe models currently in use and projected to be used in the future. Solidworks was used to conduct FEA analysis of the racks carrying a full load of pipe while being transported via forklift. Various scenarios were considered such as sudden braking, sudden acceleration, running over a bump, and impact with a structure. The final designed structure had a quoted fabrication cost of approximately $35,000 (not including all hardware). 

After the fabricated racks, machined parts, and hardware arrived, the next step was to verify the accuracy of the as-built dimensions. There were some dimensions which varied from the drawing specifications, but these variances were deemed acceptable.  

Shortly after the racks were fully assembled, I spoke often with R&D personnel to get feedback on topics such as ease of use. Some small improvements were made based on feedback. Improvements such as adhesive rulers added to the sides for easy measurement of pipe lengths, and white erasable magnetic strips added to the front restraint bars for easy pipe labelling.