Milestone 7 Paper
Advanced Portable Force Plate (APFP)
Team: Dana Lorenz, JP Ruzbasan, Sam Bianco
Advisor: Dr. Arico and Dr. Asaki
Sponsor: Dr. Solomito at Connecticut Children’s Medical Center (CCMC)
Table of Contents
Introduction 2
Project Timeline Update 2
Testing- Data Analysis 4
Design Modification 8
Final Parts List/Final Prototype Budget 9
Conclusion 11
1
Introduction
The goal of this project is to create an affordable force plate that is capable of tracking a
patient's center of pressure (COP) to establish quantitative balance data post-concussion for
the Connecticut Children's Medical Center (CCMC). A list of project objectives was created to
ensure that the remaining objectives were completed in a timely fashion. Some additional
testing and analysis of the device was completed to help aid in material decisions. Another
static simulation was completed in SolidWorks on the top piece of the force plate. The
SolidWorks simulation showed the theoretical minimum and maximum displacement for the
piece as weight was added to it. This was done to determine if the thickness of MDF wood, with
ABS plastic attached, selected was rigid enough to withstand the weight being applied to the
plate. Testing was completed on the force plate itself to ensure that the program accurately
tracks the patient’s COP. The stepper motor used for the perturbation mechanism was also
tested to ensure that it rotated the intended amount. The final design modifications were also
made at this point as well as material and budget decisions.
Project Timeline Update
A list of Project Objectives, with specific descriptions and measurable goals was created
along with tentative due dates for all objectives. The updated project objectives table can be
seen below in Table 1. The timeline needed to be altered slightly due to a change in work flow.
The final presentation was set on May 8th. While working toward the final presentation, the
updated parts list was completed later than expected on April 30th. The prototype budget was
based heavily on the updated parts list; this resulted in it being completed later as well, on May
1st.
Milestone Content Time-Depe
Specific Measurable
ndent
After the fall semester there
Each team member will
were several changes that
have recapped one or
need to be made. It is
more of the fall
important to recap all of the
milestones. Each recap
decisions and work that was
Fall Semester Recap will be approximately one 2/2/2018
completed. Each milestone
paragraph long and
should be summarized with a
include the most
focus on the most important
important parts of the
decisions that were made
milestone
during that milestone.
Looking at the new design,
All possible failure modes
what are ways that the device
are considered and
FMEA 2.0 can fail, how can they be 2/13/2018
possible solutions are
prevented by altering the
determined
device's design
2
A step by step list of
Editing the testing protocol physical and software
Testing - Protocol
from last semester for actions to be used to test 2/17/2018
Development
effectiveness and user needs the design and build
quality of the force plate
The questionnaire is
The questionnaire and
completed and specific
explanation of the
End user feedback - questions are asked in order to
questions asked will be 2/15/2018
Protocol get the most out of the
completed and formatted
feedback from the
for the paper
practitioners
The FMEA table, new
The updated FMEA table and
M5 testing development and
new testing protocol will be 2/20/2018
Presentation/Paper fall milestones will be
displayed and explained
finished and explained
Using SolidWorks simulations The simulation data will
are ran on the drawings of the be gathered and analyzed
Engineering Analysis design to determine if the to determine if changes to 3/24/2018
device can withstand the the parts/materials need
necessary loading to be made
Distribute the User Feedback
All User Feedback
Questionnaire to the project
End user feedback - Questionnaires have been
sponsor and other potential
Evaluation and Data returned, looked at, and 3/9/2018
users to obtain their feedback
Collection then used to make
on the prototype and then
changed to the device
begin analyzing the data
Following the testing protocol
There is usable data being
Testing - Data developed in M5, testing the
outputted by the device 3/26/2018
Collection device and its components to
that can be analyzed
begin obtaining data
The engineering analysis, end The engineering analysis,
user feedback analysis, and end user feedback
M6
data collection phases of the analysis, and data 3/27/2018
Presentation/Paper
project are shown and collection phases of the
explained project are completed
The data from the collection
Testing - Data phase is analyzed and used to All data collected is
4/15/2018
Analysis determine if modifications analyzed
need to be made to the device
Based on the results from the Design modifications are
Design Modification data analysis portion, design made based on the data 4/20/2018
modification will be analysis and data
3
determined to obtain more collection and analysis are
reliable data then done on the newly
design device
A complete list of
Determine the components of necessary parts and
Updated/Final Parts 4/22/2018
the device and create a list of components for
List 4/30/2018
all parts construction, including 3D
printed parts
Table outlining each piece
of the prototype, where it
Create a budget and total cost
was purchased, how 4/23/2018
Prototype Budget of the final device based on
much was purchased, and 5/1/2018
the parts list
finally the total amount
spent on that piece
The data analysis, design The data analysis, design
M7 modifications, final parts list, modifications, final parts 5/1/2018
Presentation/Paper and budget are shown and list, and budget are 5/8/2018
explained completed
Table 1: List of Project Objectives
Testing – Data Analysis
Testing is an important part of the design process. It allows for the various parts of the
device to be evaluated and calibrated before initial use. This is important because if a
component is not functioning properly or is faulty, it can be noticed early and prevent failure or
injury. Testing has been done on all of the aspects of the design. The top piece of the design
was tested using simulations to determine what material would be best to use, that will
minimize deformation. After creating a final assembly of the parts and applying a material to
them, different forces were applied to the top of the piece. The minimum and maximum
deflection of the part was then observed and recorded. The final version of the force plate top
piece was tested between weights of 50 and 300 pounds, and the results were compared to the
results from the previous design of the force plate. The sensors were tested individually,
previously in Milestone 6, as well as in the force plate. The perturbation motor was also tested
to ensure that it can move the force plate accurately when there is weight on the plate.
Below, in Figure 1, is the original static force analysis of the top piece of the force plate,
performed during milestone 6. This analysis is looking at the deformation of a piece of ABS
plastic while under 25 pounds of force. To the right of the model is a legend showing a range of
colors, each relating to a different level of deformation.
4
Figure 1: Initial static force analysis of force plate top piece
Figure 2, below, is the final static force analysis of the top piece of the force plate. This
analysis is looking at the deformation of the combined ABS plastic and MDF top pieces under 50
pounds of force. Similar to the initial force testing, the legend to the right of the model has a
range of colors, each corresponding to a level of deformation.
Figure 2: Final static force analysis of force plate top piece
5
The initial analysis had problems, particularly with the deformation showing in areas
that did not make sense. Additionally, only one piece was being evaluated at a single time. A
simpler method of testing was used which showed improvements in the visual representation
of the deformation of the top ABS plastic and MDF pieces. Having both pieces be tested in a
single analysis showed improved deformation accuracy for the entire piece when compared to
the hand calculations. Below is a table showing the force and maximum deflection of the final
force analysis under a force of 300 pounds. This table confirms the strength of the materials.
Table 2: Values of maximum deformation (mm) and applied force (lb.) from the final static force analysis
These values are obtained from running the test again across the 0-300 lbs. force in 50 lbs.
increments. Figure 2 is a visual representation of how the MDF and ABS pieces will deform; this
deformation would be exaggerated under the 300 lbs. force that is calculated in the table.
When comparing the two static force analyses done in SolidWorks, a considerable
improvement is shown form the first to the second. This is due to a better understanding of
how to use the static force analysis tools as well as a better assembly of the force plate.
Additionally, testing was completed on the sensors to ensure that all sensors work
properly individually. Once it was known that all sensors worked; known weights were applied
to them and the corresponding voltages were recorded. From this data, an equation for each
sensor was found that could convert the voltage from the sensor to a weight. This calibration
method was completed twice for each sensor. The analysis for the sensors 0-2 can be seen in
Appendix D. There were some issues with sensor 3 (S3) reading properly. To troubleshoot this
issue, the analog input of NI USB-6009 was tested along with the amplifier circuit that the
sensor was connected to. It was determined that neither of these factors were causing the
issue with S3, alternative sensors were then ordered but had similar problems as the original
sensor. It was determined that a faulty alligator clip that was being used to connect S3 to the
circuit was the problem. The sensor was then recalibrated using the previously mentioned
method. The analysis for the new sensor can be seen below in Figure 3.
6
Figure 3: Voltage vs. Weight graph of Sensor 3 calibration/testing trials
Once the testing and calibration of the individual sensors was completed, the sensors
were then put into the force plate configuration. The force plate configuration can be seen
below in Figure 4.
Figure 4: Top-down view of the force plate and sensor configuration
After the force plate was set up, the next step was to test the force plate/code. This is done to
ensure that the force plate is reading weights properly as well as tracking the COP of someone
moving while standing on it. To test this, a subject will stand on the force plate and shift their
weight; forward/backward and left/right, a different person watches the user interface to
determine if the COP indicator, on the stabilogram, is moving with the subject. An example of a
stabilogram that appears on the user interface can be seen below in Figure 5.
7
Figure 5: Patient’s stabilogram showing motion of the COP
Testing was also completed on the stepper motor that is used to create the
perturbations. The below calculations were tested to check the program that was created;
converting the user input to pulses via the equations is the most likely area of potential error.
This conversion was done using the equations seen below;
To verify that the conversion was completed properly a flag was attached to the motor shaft
and different degrees were inputted into the perturbation code. Then using a goniometer, the
degrees rotated were tracked by setting the start point to be 0° and the end point to be the
user inputted angle. The testing setup can be seen below in Figure 6.
8
Figure 6: Experimental setup testing the perturbation code at 30°
Design Modification
The final design, Figure 7, included many changes to both the physical design and coding
of the force plate. The force plate portion of the device was modified to allow for a larger
weight range that can accurately use the device and more accurate COP tracking. In the initial
design the sensors used were just for a proof of concept and could not read large enough
forces. This resulted in the interface for the force sensors and the code for the force plate to
needing to be changed. The largest physical modification that was made from the initial design,
seen in Milestone 4, is the method used to create the perturbations. Due to these changes and
the need to reduce the overall weight of the device, some of the original material decisions also
needed to be changed. These design modifications were necessary to create a device that met
all of the necessary requirements while still remaining within an acceptable budget.
Figure 7: Solidworks assembly of final force plate design
The initial prototype that was constructed in Milestone 4 used two 25 lb. FlexiForce
sensors and interfaced with LabVIEW via an Arduino. However, the final prototype needed to
function with four sensors as well as be able to read a larger weight range. In order to read the
larger weight range four, 50 kg (110 lb.) Button Load cells were used in place of the FlexiForce
sensors. This allowed for the weights that can accurately be read by the force plate to be
between about 40 lbs. and 300lbs. Due to the change in sensors, the interface between the
9
sensors and computer needed to be changed, the new interface is a NI USB-6009. This allows
for an easier interface between the sensors and LabVIEW. The LabVIEW code for the force plate
was modified to read four sensors and calculate the COP between the four sensors. The
(F −F )+(F −F )
x-position of the COP was calculated using x = X2 * [ F2 +F3 +F 0+F 1 ] , where X = 20.5 , and the
1 2 3 4
Y (F +F )−(F +F )
y-position was calculated using y = 2 *[ F2 +F3 +F 0+F 1 ] ,
where Y = 12.5 . Also, by request of
1 2 3 4
the project sponsor, the force plate program can also record the COP data points and calculate
the area of the movement of the subject on the force plate. The area of movement was
calculated using A = abπ , where a and b are the distances from the center of the area covered
in the x and y directions respectively. The block diagram and front panel for the force plate
portion of the device can be seen in Figure F1 and Figure F2.
The method of perturbation had to be changed in order to create a perturbation
mechanism able to move the necessary weight. The new method of perturbation involves a
central stepper motor that rotates the top portion of the device, the force plate, in a clockwise
or counter-clockwise direction. The stepper motor was selected after considering the amount
of torque that would be needed. The control of the motor was twofold; the immediate
interface was the stepper motor driver which was attached to the motor by the four wires of
the motor. The second control method was a combination of the NI USB-6009 devices. They
created the pulses that would go to the motor driver and in turn cause the motor to rotate as
well as the signals for enable and direction. All of these connections were controlled via one
perturbation code. Because this iteration of the device used the NI USB-6009 devices, the code
shifted away from MATLAB towards LabVIEW. Executable files would be a sufficient method of
delivery for the project sponsor. In the perturbation code, every event that took place would
need to happen in a specific order. The documentation for the motor driver stated that certain
inputs needed to be set before others. In order to do this the correct way, one of the USB-6009
devices was responsible for the enabling the driver and controlling the direction that the motor
would rotate, while a second was responsible for the pulses. Use of two USB-6009 devices has
never been done before in this capacity; therefore, the literature on the control and correct
programming does not exist. The team had to troubleshoot every issue and create solutions
from scratch. Controlling the motor driver was done by the specifications listed in the motor
driver’s manual. Each of the inputs was controlled by a 5V signal; logic high was 5V and logic
low was 0V. Enabling the driver required a logic high signal; the rotation direction in
comparison used logic high to be counter clockwise and logic low to be clockwise. This
transitioned into the pulse section of the code where the user will input degrees would be
converted into pulses and rounded to the nearest integer to ensure the proper control of the
motor. Non-integer inputs trying to be output as pulses would result in an incorrect rotation of
the motor or no motion at all. Final portions of the code reverse the motor direction logic that
was set at the beginning of the code. After the “Return” button is pressed the motor will move
the same number of degrees as set previously, but in the opposite direction. To end the code
10
completely, all outputs are set to logic low, preventing any unwanted movements or stagnant
voltages being sent to the motor driver from the USB-6009 devices.
Important material decisions were made regarding the top portion and perturbation
housing. The top portion, which includes the rigid top plate and sensor housing, was made
lighter and stronger. Using a combination of a piece of ABS plastic on top of a rigid MDF board
allowed the top portion of the force plate to be both strong and light when compared to
previous iterations. The sensor housing, which sits below the rigid ABS and MDF boards, is
made from high density polyethylene (HDPE) plastic and allows the force plate to be used by
itself without the perturbation housing. The perturbation housing was redesigned to be slightly
smaller in width and length. Additionally, the material used to create the perturbation housing
was changed from wood to HDPE to reduce the weight of the device.
Final Parts List/Final Prototype Budget
The final parts list is a combination of parts that had been purchased by the team and
parts that were given to the team by the project sponsor or the University of Hartford. If the
device was to be manufactured from scratch the cost is calculated in the table below. The total
cost for the raw materials and electronics needed was $785.00; this excludes all potential costs
that would be added due to tax or costs added due to the physical creation of the device. The
sheet of HDPE is used in the fabrication of the bottom and sides of Figure E2 as well as Figure
E4. The force plate is made from the ABS sheet, MDF sheet, and force sensors; in order to get
the correct readings from the force sensors, one of the NI USB-6009 devices is connected to the
force sensor circuit. Creating the circuit required the INA 122 chips along with the resistors and
capacitors. It was seated in the Perfboard and soldered in place. The second NI USB-6009 was
used in the bottom of the device where the perturbations were controlled. Using the NI
USB-6009’s ability to send pulses, the stepper motor was used effectively. The pulses went to
the stepper motor driver and then to the stepper motor itself. Utilizing pulses allows the user to
rotate the motor at specific angles. Due to the higher torque of the motor, a housing needed to
be designed and then was 3D printed. It included a body housing and a cap that fit snugly on
the top. The cap bolted on to the motor and the entire 3D printed casing bolted down to the
bottom of the device, Figure E4. The last 3D printed part was the cross-shaped connection
piece, Figure E7. The stepper motor shaft would sit in the center while the cross allowed for the
top force plate to sit on three bolts that were through the holes on the cross. Supporting all of
the weight from the force plate rests on the ball wheels and their risers. The ball wheels allow
for uninhibited force plate rotation. An important part of the uninhibited rotation is the height
of the risers for the ball wheels. The risers are cut to a specific length that accounts for the
space needed to keep the motor from bearing any weight. When set correctly the perturbation
mechanism works in unison to rotate the top of the force plate to the desired angle.
11
Part Name Detailed Description Quantity Cost x1 Total Cost
24" by 48" by 0.25" thick sheet of
HDPE sheet HDPE for the bottom perturbation 1 28.64 28.64
enclosure and the sensor housing
24" by 24" by 0.0625" thick sheet of
ABS sheet ABS for the smooth surface that 1 13.32 13.32
patients will stand on
Sheet of MDF that serves as the
rigid piece of the force plate. It will
MDF sheet have the force sensors attached to 1 5.68 5.68
it as well as the force sensor circuit
and NI USB-6009
50 kg button load cells that have an
Force sensors excitation voltage of 5V with a 4 45.00 180.00
rated output of 1mv/V
Data acquisition device with analog
inputs and digital outputs that can
NI USB-6009 be used for the force sensors as 2 200.00 400.00
well as sending pulses to the motor
driver
Amplifier chip that does not need a
negative power supply unit in order
INA 122 chip 4 3.58 14.32
to work. Allows for a multitude of
different gain values
Four 200-ohm resistors and four 0.1
Resistors/ capacitors 1 3.00 3.00
micro farad capacitors
Adafruit Perma-Proto Half-sized
Perfboard 1 4.50 4.50
Breadboard PCB - Single
Wires of varying length that were
needed to connect the circuits to
Assorted wires 1 4.00 4.00
the NI USB-6009 as well as to the
motor driver and power supply
Stepper motor with L=39mm Gear
Nema 17 Stepper Motor Ratio 15:1 High Precision Planetary 1 67.29 67.29
Gearbox
Stepper motor drive with ratings of
0.3-2.2A 18-30VDC. Designed for
Digital Stepper Driver 1 22.35 22.35
Nema 8, 11, 14, 16, 17 Stepper
Motor
Ball bearing-like wheel to be
mounted in the corners of the
Ball wheel 4 5.02 20.08
bottom half of the device to allow
for the top half to move freely
2" by 4" pine risers used under the
Wood riser 1 3.19 3.19
ball wheels
12
Assorted 3D printed parts including
the perturbation connection piece
3D printed parts 1 10.63 10.63
as well as the top and bottom parts
to the stepper motor housing
Screws, nuts, and bolts that were
Assorted hardware 1 8.00 8.00
used in the project
Total 785.00
Conclusion
Throughout this final milestone, a functional prototype was created. In order to
complete the project on time some of the dates in the timeline needed to be adjusted. More
time was allocated for the device construction; this pushed back the final parts list and budget
of those parts until the final construction had been completed. This device is able to track a
patients COP as it moves in real time; the COP data points are tracked and then used in
calculating the area of movement. This provides a simple number for the evaluating doctors to
look at when analyzing a patient’s balance. Before the functional prototype was finished,
testing of each component as well as static simulations were done to ensure the device would
work. Static testing used the materials that had been decided on by the team; a goal with the
material decisions was to cut down on the overall weight of the device while maintaining full
functionality and weight bearing ability. Overall the perturbation mechanism that had been
chosen was powerful enough to rotate a patient but, it was not fast enough for a “jolting”
movement. Future iterations of the device can improve upon the perturbation speed and
potentially ease of carrying.
13
