← Back to all projects

PVC Knee Scooter

Mechanical design-build | ME 328 | Cal Poly San Luis Obispo

Mechanical Design Hand Calculations Finite Element Analysis Prototype Fabrication Physical Validation
Completed white PVC knee scooter outdoors, showing the padded knee platform, steering column, and wheel layout

Working with Daniel Mendes, Chananeth Manava, and Hong Dang, I designed, analyzed, built, and tested a low-cost knee scooter from Schedule 40 PVC. The project moved from anthropometric requirements and hand calculations through two FEA iterations, physical fabrication, and test-based validation.

01

Design requirements and architecture

A stable, low-cost alternative to crutches

The design supports an injured leg on a padded platform while the rider pushes with their healthy leg and steers through a rotating front column. A wide front track improves lateral stability, while a compact rear wheel assembly leaves clearance for the rider's pushing leg.

The final configuration uses a rectangular, truss-like PVC frame to distribute knee loads and limit bending. Cross fittings support the knee platform, and bored end caps create a pivot for the steering column. The geometry was sized from anthropometric measurements rather than selected by eye.

$173.73 Reported prototype material cost
3 cases Normal, backpack, and handlebar loads
1 & 1.5 in Schedule 40 PVC sizes used
Dimensioned hand sketch of the selected knee scooter configuration
The selected concept established the wheel layout, frame proportions, and knee-pad location.
CAD model of the PVC knee scooter frame and steering column
The CAD model translated the hand sketch into a geometry suitable for structural analysis.
Design decision: Two earlier concepts were rejected because they required custom parts and more complex fabrication. The selected PVC architecture could be built with readily available fittings and student-shop tools.
02

Load cases and hand calculations

Bounding normal use and foreseeable misuse

I evaluated three scenarios: the rider's weight on the knee platform, the rider plus a backpack, and an abuse case with a 10 lb bag hanging from each handlebar end. The heavier knee-platform case governed the frame calculation, while the bag load isolated bending and deflection in the handlebars.

With a safety factor of two applied, the knee support was checked under a 246 lb design load. Simplified beam theory predicted 0.080 in of platform deflection. For the handlebar case, a 90 lb-in bending moment produced an estimated 262 psi maximum stress and 0.013 in tip deflection.

246 lb Factored knee-platform design load
0.080 in Calculated knee-support deflection
262 psi Calculated handlebar bending stress
Load case Purpose Applied condition Primary check
Normal use Daily operation User load on knee pad Frame stiffness
User + backpack Maximum expected use 246 lb factored design load Knee-platform deflection
Handlebar abuse Foreseeable misuse 10 lb at each handlebar end Steering-column bending
03

FEA-driven optimization

Reducing excess strength without sacrificing safety

The first solid-element model used 1.5 in PVC throughout, fixed the four wheel-mount locations, and applied 150 lb over the knee-support area plus 10 lb at each handlebar end. Its maximum stress was approximately 0.65 ksi, corresponding to a safety factor near 10 against the reported 6-8 ksi ultimate strength of PVC.

Because the first design carried more margin than necessary, the steering column and handlebars were reduced to 1 in PVC. The final model predicted approximately 0.905 ksi maximum stress and 0.149 in maximum displacement, with the most critical region located where the steering column meets the front crossbar.

0.905 ksi Final FEA maximum stress
0.149 in Final predicted maximum displacement
6-7 Final simulated safety-factor range
Stress and displacement results for the initial all-1.5-inch PVC scooter model
The initial 1.5 in model produced a safety factor near 10, indicating room to reduce material.
Stress and displacement results for the final scooter model with a 1-inch steering column
The final iteration used a smaller steering column and retained a safety factor between 6 and 7.
Key lesson: The analysis was not only a pass/fail check. Comparing the model with hand calculations gave enough confidence to remove unnecessary material while preserving a conservative margin for a brittle structure.
04

Fabrication and prototype testing

Turning a simulation model into a working mobility device

The frame was dry-fit first so dimensions, alignment, and wheel placement could be corrected before permanent bonding. After verification, the structural joints were primed and glued, the steering column was left free to rotate, threaded rod formed the wheel axles, and foam over a wood platform created the knee support.

Testing covered static load capacity, handlebar loading, structural integrity, turning, and rider fit. The unglued frame supported roughly twice the intended weight with little visible deflection. A later 20 lb handlebar load produced about 0.0625 in of measured deflection, and no measurable main-frame deflection was observed at the intended user load.

~2x Intended load supported in testing
20 lb Total handlebar abuse load
0.0625 in Measured handlebar deflection
Completed PVC knee scooter standing outdoors
Completed prototype showing the wide wheelbase and padded knee platform.
Close view of the finished PVC scooter frame and steering column
Close-up of the glued frame joints, steering column, and wheel mounting.
Side view of the PVC knee scooter carrying a backpack from the handlebars
A loaded backpack recreated the handlebar abuse case during physical testing.
Rear view of the loaded PVC knee scooter
The rear view shows rider clearance and the backpack load path into the steering assembly.
05

Validation and next iteration

Comparing prediction, observation, and user feedback

Both FEA and physical testing identified the handlebar ends as the most flexible region. The measured 0.0625 in displacement was less than half of the 0.149 in prediction, so the model was conservative while still capturing the correct deformation pattern. The extra stiffness likely came from glued fittings and local reinforcement that were simplified in the simulation.

Physical testing also exposed issues the structural model could not: dry-fit joints twisted during sharp turns, fitting stack-up made the scooter too tall for the original user, and the steering-column connection remained the main source of wobble. A future version should add adjustable handlebar and knee-platform heights, reinforce the steering pivot, use aluminum tubing for durability, and include a hand-operated brake.

0.149 in FEA-predicted handlebar displacement
0.0625 in Physically measured displacement
< 50% Measured result relative to prediction
Engineering takeaway: Hand calculations bounded the problem, FEA revealed the critical regions, and prototype testing validated the deformation pattern while uncovering assembly and ergonomic issues that the model could not represent.