Dataset: Modeling manual wheelchair propulsion cost during straight and curvilinear trajectories Jacob Misch 1,2 Morris Huang 1,3 Stephen Sprigle 1,2,4 1 Rehabilitation Engineering and Applied Research Lab, Georgia Institute of Technology, Atlanta, Georgia, United States of America 2 School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States of America 3 Department of Bioengineering, University of Colorado Denver, Denver, Colorado, United States of America 4 School of Industrial Design, Georgia Institute of Technology, Atlanta, Georgia, United States of America Email contact: rearlab@gatech.edu This work was supported by internal funding from the Rehabilitation Engineering and Applied Research (REAR) Lab. This data was used in the manuscript submitted and accepted for publication in the Public Library of Science (PLOS) ONE journal, with manuscript number: PONE-D-20-01476R1 Title: Modeling manual wheelchair propulsion cost during straight and curvilinear trajectories Authors: Jacob Misch, Morris Huang, Stephen Sprigle The methods of data collection for the energy loss parameters and propulsion costs are summarized in the journal article, with references to more detailed articles. The data analysis methods are described in depth within the journal article associated with this dataset, which is open access. Data was collected between November 2016 to August 2018. This Excel workbook includes a reference worksheet with geometric notation and three worksheets with data: "Wheelchair Geometry" - A graphical depiction of the geometric measurements of the wheelchair frame used in the System Loss equations "Straight" - Data related to the "Straight" maneuver including rolling resistance, scrub torque, and the acceleration/steady-state propulsion costs. The propulsion cost is the main outcome variable of the study. Data descriptors can be found in the worksheet. "Fixed-Wheel" - Data related to the "Fixed-Wheel Turn" maneuver including rolling resistance, scrub torque, and the acceleration/steady-state propulsion costs. The propulsion cost is the main outcome variable of the study. Data descriptors can be found in the worksheet. "Alter. Zero-Radius" - Data related to the "Alternating Zero-Radius Turn" maneuver including rolling resistance, scrub torque, and the acceleration/steady-state propulsion costs. The propulsion cost is the main outcome variable of the study. Data descriptors can be found in the worksheet. "Straight" columns: Surface - Either "Tile" or "Carpet". The AMPS and the component tests were all conducted with the identified surface in each row. Drive_Wheel - The name of each tested drive wheel. "Solid Mag" is a 24"-diameter solid polyurethane tire. "Spinergy" is a 24"x1" Schwalbe Spinergy wheel with Schwalbe Marathon Plus tires (100 psi inflation). "Spoked" is a 24" x 1-3/8" Primo Orion Standard Pneumatic (75 psi inflation). Caster - The name of each tested caster. WD_Config - The weight distribution of the system, represented by the percentage of total system weight over the drive wheels. Total_Cost - The propulsion cost for the entire maneuver, comprised of acceleration, steady-state, and deceleration. This column is exclusively used for the Alternating Zero-Radius Turn manevuer, and unused for the Straight and Fixed-Wheel Turn maneuvers. Main_Cost - The propulsion cost for the combined acceleration and steady-state phases of the Straight and Fixed-Wheel Turn maneuvers. This column is unused in the Alternating Zero-Radius Turn maneuver. Accel_Cost - The propulsion cost for the acceleration phase of the Straight and Fixed-Wheel Turn maneuvers. This column is unused in the Alternating Zero-Radius Turn maneuver. Sys_Mass - The combined mass of the wheelchair (~13 kg) and the AMPS (80 kg or 100 kg, depending on loading conditions). Sys_Yaw - The yaw moment of inertia about the vertical axis of the system, located at the center of mass. This is recorded with the iMachine. DW_Inertia - The pitch moment of inertia of a single drive wheel about the axis passing through the axle bearing. This is measured with a trifilar pendulum system. Caster_Inertia - The pitch moment of inertia of a single caster about the axis passing through the axle bearing. This is measured with a trifilar pendulum system. WD_Value - The actual weight distribution of the system, as measured by the iMachine. Caster_Scrub - The resistive scrub torque, in Newton-meters, of the specified surface against the caster after 'conditioning' trials to remove the material sheen. DW_Scrub - The resistive scrub torque, in Newton-meters, of the specified surface against the drive wheel after 'conditioning' trials to remove the material sheen. Caster_RR - The rolling resistance value of the specified caster, in Newtons, as determined by the coast-down cart protocol. DW_RR - The rolling resistance value of the specified drive wheel, in Newtons, as determined by the coast-down cart protocol. Sys_Loss_Value - A term representing the major sources of energy loss during the maneuver, calculated using the equation: [2 * Caster_RR] + [2 * DW_RR] "Fixed-Wheel" columns: Surface - Either "Tile" or "Carpet". The AMPS and the component tests were all conducted with the identified surface in each row. Drive_Wheel - The name of each tested drive wheel. "Solid Mag" is a 24"-diameter solid polyurethane tire. "Spinergy" is a 24"x1" Schwalbe Spinergy wheel with Schwalbe Marathon Plus tires (100 psi inflation). "Spoked" is a 24" x 1-3/8" Primo Orion Standard Pneumatic (75 psi inflation). Caster - The name of each tested caster. WD_Config - The weight distribution of the system, represented by the percentage of total system weight over the drive wheels. Total_Cost - The propulsion cost for the entire maneuver, comprised of acceleration, steady-state, and deceleration. This column is exclusively used for the Alternating Zero-Radius Turn manevuer, and unused for the Straight and Fixed-Wheel Turn maneuvers. Main_Cost - The propulsion cost for the combined acceleration and steady-state phases of the Straight and Fixed-Wheel Turn maneuvers. This column is unused in the Alternating Zero-Radius Turn maneuver. Accel_Cost - The propulsion cost for the acceleration phase of the Straight and Fixed-Wheel Turn maneuvers. This column is unused in the Alternating Zero-Radius Turn maneuver. Sys_Mass - The combined mass of the wheelchair (~13 kg) and the AMPS (80 kg or 100 kg, depending on loading conditions). Sys_Yaw - The yaw moment of inertia about the vertical axis of the system, located at the center of mass. This is recorded with the iMachine. DW_Inertia - The pitch moment of inertia of a single drive wheel about the axis passing through the axle bearing. This is measured with a trifilar pendulum system. Caster_Inertia - The pitch moment of inertia of a single caster about the axis passing through the axle bearing. This is measured with a trifilar pendulum system. WD_Value - The actual weight distribution of the system, as measured by the iMachine. Caster_Scrub - The resistive scrub torque, in Newton-meters, of the specified surface against the caster after 'conditioning' trials to remove the material sheen. DW_Scrub - The resistive scrub torque, in Newton-meters, of the specified surface against the drive wheel after 'conditioning' trials to remove the material sheen. Caster_RR - The rolling resistance value of the specified caster, in Newtons, as determined by the coast-down cart protocol. DW_RR - The rolling resistance value of the specified drive wheel, in Newtons, as determined by the coast-down cart protocol. Sys_Loss_Value - A term representing the major sources of energy loss during the maneuver, calculated using the equation: [DW_Scrub] + [s * DW_RR] + [Caster_RR * √( ( ( s - p )/2 )^2 + d^2 )] + [Caster_RR * √( ( ( s + p ) / 2 )^2 + d^2 )] "Alter. Zero-Radius" columns: Surface - Either "Tile" or "Carpet". The AMPS and the component tests were all conducted with the identified surface in each row. Drive_Wheel - The name of each tested drive wheel. "Solid Mag" is a 24"-diameter solid polyurethane tire. "Spinergy" is a 24"x1" Schwalbe Spinergy wheel with Schwalbe Marathon Plus tires (100 psi inflation). "Spoked" is a 24" x 1-3/8" Primo Orion Standard Pneumatic (75 psi inflation). Caster - The name of each tested caster. WD_Config - The weight distribution of the system, represented by the percentage of total system weight over the drive wheels. Total_Cost - The propulsion cost for the entire maneuver, comprised of acceleration, steady-state, and deceleration. This column is exclusively used for the Alternating Zero-Radius Turn manevuer, and unused for the Straight and Fixed-Wheel Turn maneuvers. Main_Cost - The propulsion cost for the combined acceleration and steady-state phases of the Straight and Fixed-Wheel Turn maneuvers. This column is unused in the Alternating Zero-Radius Turn maneuver. Accel_Cost - The propulsion cost for the acceleration phase of the Straight and Fixed-Wheel Turn maneuvers. This column is unused in the Alternating Zero-Radius Turn maneuver. Sys_Mass - The combined mass of the wheelchair (~13 kg) and the AMPS (80 kg or 100 kg, depending on loading conditions). Sys_Yaw - The yaw moment of inertia about the vertical axis of the system, located at the center of mass. This is recorded with the iMachine. DW_Inertia - The pitch moment of inertia of a single drive wheel about the axis passing through the axle bearing. This is measured with a trifilar pendulum system. Caster_Inertia - The pitch moment of inertia of a single caster about the axis passing through the axle bearing. This is measured with a trifilar pendulum system. WD_Value - The actual weight distribution of the system, as measured by the iMachine. Caster_Scrub - The resistive scrub torque, in Newton-meters, of the specified surface against the caster after 'conditioning' trials to remove the material sheen. DW_Scrub - The resistive scrub torque, in Newton-meters, of the specified surface against the drive wheel after 'conditioning' trials to remove the material sheen. Caster_RR - The rolling resistance value of the specified caster, in Newtons, as determined by the coast-down cart protocol. DW_RR - The rolling resistance value of the specified drive wheel, in Newtons, as determined by the coast-down cart protocol. Sys_Loss_Value - A term representing the major sources of energy loss during the maneuver, calculated using the equation: [2 * Caster_Scrub] + [2 * DW_RR * ( s / 2 )] + [2 * ( Caster_RR * √( ( p/2 )^2 + d^2 ) )] Methods: The objective of this study was to combine the results from empirical component-level and system-level tests to generate statistical models describing the relationships between inertial or energy-loss parameters to the mechanical propulsion cost across different wheelchair configurations during straight and curvilinear trajectories. The empirical measurements utilized several protocols. Inertial parameters (system mass, yaw inertia, and weight distribution) were measured with the iMachine, an inertial measurement tool consisting of an oscillating platform described in [1]. Energy loss sources (or non-conservative forces and torques acting on the wheelchair) were measured with two separate techniques. Rolling resistance forces were collected using an instrumented coast-down cart. Wheel-mounted accelerometers recorded the cart velocity and average deceleration during an analysis window between 0.95 m/s to 0.65 m/s. The cart was loaded with metal bar weights to apply loading conditions to the wheels similar to those of an occupied wheelchair [2]. Resistive scrub torques were measured with a modified material-testing load frame by recording the torques required to rotate a weighted scrub plate against each wheel. The scrub plates used samples of low-pile carpet and linoleum tile to investigate the real-world terrain traveled by wheelchair users [2]. Finally, the Anatomical Model Propulsion System (AMPS) wheelchair-propelling robot [3] was used to measure the mechanical propulsion cost of over-ground travel during straight and curvilinear trajectories [4]. For this dataset, the AMPS was placed onto an ultralightweight wheelchair (Quickie GT) and trials were run across numerous configurations, spanning three different drive wheels, four different casters, two occupant masses (AMPS masses of 80 kg and 100 kg), and three weight distributions (60%, 70%, 80% mass over the rear axle) over two surfaces (tile, carpet). Three distinct maneuvers were executed: straight motion, fixed-wheel turning, and alternating zero-radius turning [4]. Ten trials were run for each maneuver per unique configuration, resulting in 4,280 total over-ground trials. General linear models were used to investigate the relationships between the propulsion cost and the other measurements. These models were unique to each maneuver as each was given distinct predictor variables, though they all shared the single common outcome variable of propulsion cost. The initial model for the straight maneuver was given predictor variables of system mass, weight distribution, and a system loss variable that was a function of the caster and drive wheel rolling resistances. The model identified the system loss term as the sole predictor, so the other variables were removed and a subsequent model was generated with only system loss. The model for the fixed-wheel turn maneuver was given yaw inertia, weight distribution, mass, and a system loss variable that was a function of caster rolling resistance, drive wheel rolling resistance, geometric measurements of the wheelchair frame, and the drive wheel scrub torque. The initial model identified the system loss parameter as the sole predictor. The second fixed-wheel turn model used system loss as the only predictor variable. Finally, the initial model for the zero-radius turn maneuver was given the same parameters as the initial fixed-wheel turn model, except that the system loss variable for zero-radius turning removed drive wheel scrub torque and included caster scrub torque instead. This model identified system yaw and system loss as the two significant predictor variables. The subsequent model was run using only these two predictors. (Detailed methods can be found in the PLOS ONE article, which includes references to instrumentation and more detailed articles on the equipment used.) References in Methods: [1] Eicholtz, M., et. al, J Rehabil Res Dev, 2012, 49(1), 51-62. [2] Sprigle, S., et. al, Assist Technol, 2019, 1-13. [3] Liles, H., et. al, IEEE Trans Neural Syst Rehabil Eng, 2015, 23(6), 983-991. [4] Sprigle, S., et. al, J Rehabil Assist Technol Eng, 2020, 7, 1-14.