ME 321 Mechanical Engineering Analysis for Design · Spring 2025
Pump Support Structure
A class project for a fictional client, PumpCo, who needed a rigid, adjustable platform to hold a 12.5 lb centrifugal pump driven at 1800 RPM by a belt with 200 N of tension. The structure had to bolt into an existing frame via clevis pins, survive a wet environment, and stay under 1 mm of tip deflection at a factor of safety of at least 1.5.
Team of three (Ethan Fernandez, Pierre Gathy, Connor Mitchell). Analysis, CAD, and fabrication were shared across the team — no individual ownership of any one part.
What it is
PumpCo, the made-up client the course gave us, wanted a frame-mounted pump platform that would not flex under torsional load from the belt drive. The pump itself runs water at 150 GPM against 60 ft of head — about 2.3 fluid horsepower, 3.5 mechanical horsepower at 65% efficiency, which translates to ~13.85 N·m of pulley torque and ~509 N of net pulley reaction. Everything downstream of those numbers is what the structure has to absorb without yielding, buckling, or deflecting more than a millimetre.
How it works
The structure is two angled strut arms in compression and two base support arms in combined bending, torsion, and axial compression, all connected to a flat aluminum pump platform with clevis pins (treated as pins for the analysis). Material is 6061 aluminum throughout, picked for the corrosion requirement and machinability against a $150 raw-stock budget.
We sized everything by hand first: free body diagrams of the platform, pulley, and assembly; full force and moment balances; symbolic stress at each location with stress-concentration factors pulled from Shigley's tables (transverse-hole-in-bending, transverse-hole-in-compression, etc.). That gave us nominal and peak stresses for every component, principal stresses via Mohr's circle, and a yielding and Euler-buckling check on every load path.
Once the symbolic work locked in the geometry, we ran ANSYS FEA on the as-modelled SolidWorks assembly to validate stresses and predict the platform tip deflection. The factors of safety were comfortably high in the strut arms (over 1400×) and lowest in the base arms (still >1.5), which matched the analytical result.
Where it ended up
Built and tested in the test frame at the end of the semester. The structure carried the full required 115 lb load and was pushed up to 200 lb without failure. Maximum deflection at the platform tip came in around 5 mm — over the 1 mm spec, and worse than the 0.067 mm the analytical model predicted and the 2.3 mm FEA predicted. That gap traces to the analytical model assuming ideal supports and ignoring fastener compliance, both of which the physical assembly clearly carried.
The other lesson from the build was a fit issue with the supplied test frame; we ended up 3D-printing custom spacers at the interface so the prototype could mount cleanly. Useful reminder that a design that closes on paper still needs to clear tolerances on whatever it's bolting into.