ME 3057 Homework 6 Force Transducer

NAME_______Repp Ledger_____________________________ SECTION_____G_____

Notes:
● Please highlight your responses questions.
● Use MATLAB to solve this homework.
● Submit any code used to generate your solutions at the end of the assignment. Please comment it
appropriately.

NOTE: Due to the nature of the destructive testing that will be conducted during the next two labs, safety glasses
are required and will be provided to you. As always, you are also required to wear long pants and close-toed
shoes.

The way materials respond to stress is important design consideration for mechanical engineers. Mechanical
systems can only perform correctly if the materials and the component design are correctly matched to the
functional loading as excess stress or strain can lead materials to fail.

For this block of labs you will be using a test rig designed for material stress-strain and failure testing to analyze
different materials. The test rigs are designed to place a test specimen under uniaxial loading and are capable
of measuring the applied load and the corresponding displacement. This week’s lab will focus on understanding
the force transducer and signal path design, calibration, and implementation as well as testing and
characterizing the performance of a climbing sling.

Figure 1: Test rig design showing orientation of force transducer and test specimen during stress/strain testing.
Test specimen is strained by use of a lead screw whose displacement is measured by a displacement sensor.

Force Transducer and Signal Path Design

1
ME 3057 Homework 6 Force Transducer
To analyze the stress on an object, it is important to know the force that is being applied. But force is hard to
measure directly in many objects. Force transducers can be used to measure force indirectly. One way to do this
is to measure the deformation of an object with a known relationship between to the force applied (for example
a spring scale). The force transducer used in the test rig consists of an Al-6061-T6 aluminum dog bone (Figure 2)
with an attached strain gauge. We will now explore the design of this force transducer and how it produces a
signal that can be read and processed by the myRIO.

Figure 2: Force transducer dog bone schematic

1. The dog bone used in the force transducer is designed such that in the constant cross-section area where
the strain gauge is attached it can be assumed that the force applied acts as uniaxial normal stress. For
the applied load, 𝐹, and the cross-sectional area, 𝐴, what is the normal stress, 𝜎?
𝐹
𝜎=
𝐴

2. What is the theoretical maximum load, 𝐹𝑚𝑎𝑥 , that can be applied to the force transducer before
damaging the transducer by yielding? Use the material properties for Aluminum 6061-T6 at
www.matweb.com. Do not use the Alcald or RSP aluminum properties. (Hint: should be between 40 and
50 kN).
𝐹𝑚𝑎𝑥 = 44.516 𝐾𝑁

𝐾𝑁
𝜎𝑦𝑖𝑒𝑙𝑑 =276 Mpa=0.276 𝑚𝑚2 𝐴 = 12.7𝑚𝑚 ∗ 12.7𝑚𝑚 = 161.29𝑚𝑚2
𝐾𝑁
𝐹𝑚𝑎𝑥 = 𝜎𝑦𝑖𝑒𝑙𝑑 ∗ 𝐴=0.276 𝑚𝑚2 ∗ 161.29𝑚𝑚2 = 44.516 𝐾𝑁
𝐹𝑚𝑎𝑥 = 44.516 𝐾𝑁

2
ME 3057 Homework 6 Force Transducer
3. It is assumed that as long as the load applied to force transducer is below the yielding point of the
material, the resulting deformation will act according to Hooke’s law of linear elasticity. What is the
resulting strain, 𝜖, in the gauge area as a function of the stress, 𝜎?

𝜎
𝜖=
𝐸
𝐸 = 68.9 GPa

𝜎 276 Mpa
𝜖= = = 0.004
𝐸 68.9 GPa

4. The strain gauge is used to measure the small strains which the force transducer experiences under load
by varying in resistance. What is the instantaneous resistance of the strain gauge, 𝑅𝐺 , as a function of
the initial unstrained resistance, 𝑅0 , the gauge factor, 𝐺𝐹, and the strain, 𝜖?
𝑅𝐺 = 𝐺𝑓 ∗ 𝜀 ∗ 𝑅0

Wheatstone bridges allow a small change in resistance to be measured relative to an initial voltage value, giving
better sensitivity than measurements try to compare large resistance values. Figure 3 shows a schematic of a
Wheatstone bridge and the actual circuity you will be using for the lab.

Figure 3: a) Wheatstone bridge schematic b) Wheatstone circuit board used in force transducer signal path.
The red circle highlights the connection terminals for the strain gauge and the blue circle the resistor
modified by the potentiometer. The resistors were chosen to be the same 120 Ω as the initial strain
gauge resistance. Note that the BNC cable polarity uses the voltage at D as positive and B as negative.

5. What is the Wheatstone output voltage, 𝑉𝑜𝑢𝑡 , as a function of the input voltage, 𝑉𝑖𝑛 , the strain gauge
resistance, 𝑅𝐺 ,the potentiometer resistance, 𝑅𝑃 , and the set Wheatstone bridge resistors, 𝑅𝑊 ?
3
ME 3057 Homework 6 Force Transducer
𝑉𝑜𝑢𝑡 = ?
𝑅2 𝑅3
𝑅1 = 𝑅𝐺 = 𝑅𝑃 = 𝑅2 𝑉𝐵 = 𝑉𝑖𝑛 𝑉𝐷 = 𝑉𝑖𝑛 𝑤ℎ𝑒𝑛 𝑉𝐵 = 𝑉𝐷
𝑅1 + 𝑅2 𝑅3 + 𝑅4
𝑅2 𝑅3
𝑉𝑖𝑛 𝑅 = 𝑉𝑖𝑛 𝑅 𝑅2 (𝑅3 + 𝑅4 ) = 𝑅3 (𝑅1 + 𝑅2 ) (𝑅1 𝑅3 ) = (𝑅2 𝑅4 ) 𝑅4 = 𝑅3
1 +𝑅2 3 +𝑅4
𝑅2 𝑅3
𝑉𝑜𝑢𝑡 = 𝑉𝑖𝑛 − 𝑉𝑖𝑛
𝑅1 + 𝑅2 𝑅3 + 𝑅4
𝑅𝑃 𝑅𝑊
𝑉𝑜𝑢𝑡 = 𝑉𝑖𝑛 (𝑅 −𝑅 )
𝐺 +𝑅𝑃 𝑊 +𝑅𝑊

The voltage change from the Wheatstone bridge is on the order of millivolts for the test rig setup. Since this on
the order of the myRIO resolution, an amplifier and low pass filter is used to boost the voltage and clean the
signal.

6. What is the voltage measured by the myRIO, 𝑉𝑅𝐼𝑂 , as a function of the amplifier gain, 𝐺𝑎𝑚𝑝 , and the
Wheatstone bridge output voltage, 𝑉𝑜𝑢𝑡 ?
𝑉𝑜𝑢𝑡
𝐺𝑎𝑚𝑝 = 𝑉𝑖𝑛 = 𝐺𝑎𝑚𝑝 ∗ 𝑉𝑜𝑢𝑡 𝑉𝑅𝐼𝑂 = 𝐺𝑎𝑚𝑝 ∗ 𝑉𝑜𝑢𝑡
𝑉𝑖𝑛

7. What is the voltage measured by the myRIO if the force transducer strains 0.1%? The strain gauge
specifications can be found in the lab resources. Assume the amplifier gain is 200, the input voltage to
the Wheatstone bridge is 3.3 V, and the Wheatstone bridge resistors have the same resistance as the
unloaded strain gauge (i.e. the bridge is initially balanced). (Hint: The voltage should be between 0.2 and
0.5 volts)
𝑉𝑅𝐼𝑂 = ?
𝑅𝐺 = 120 𝛺 𝑅3 = 𝑅4 𝜀 = .001 𝐺𝑓 = 2.070 𝑅𝐺 = 𝑅1 𝑅2 = 𝑅𝐺 + ∆𝑅
𝐺𝑎𝑚𝑝 = 200 𝑉𝑖𝑛 = 3.3𝑉 ∆𝑅 = 𝑅𝐺 ∗ 𝐺𝑓 ∗ 𝜀 = 0.2484
𝑅2 𝑅3
𝑉𝑜𝑢𝑡 = 𝑉𝑖𝑛 ( − )
𝑅1 + 𝑅2 𝑅3 + 𝑅4
𝑉𝑜𝑢𝑡 𝑅𝐺 + ∆𝑅 1 ∆𝑅
= − = = 0.00051696 𝑉
𝑉𝑖𝑛 𝑅𝐺 + 𝑅𝐺 + ∆𝑅 2 4𝑅𝐺 + 2∆𝑅
𝑉𝑅𝐼𝑂 = 𝐺𝑎𝑚𝑝 ∗ 𝑉𝑜𝑢𝑡 = 0.3411 𝑉
𝑉𝑅𝐼𝑂 = 0.3411 𝑉

4
ME 3057 Homework 6 Force Transducer

8. What is the theoretical gain (𝐹 ⁄𝑉𝑅𝐼𝑂 ) of the force transducer and signal path in N/V? (Hint: The gain
should be between 20,000 and 40,000 N/V)
𝜎
𝐺𝐹𝑇 = (𝐹 ⁄𝑉𝑅𝐼𝑂 ) 𝐹= 𝜎 = 𝐸 ∗ 𝜀 = 68.9 ∗ 109 𝑃𝑎 ∗ .001 = 68900000 𝑃𝑎
𝐴𝑟𝑒𝑎
𝐴𝑟𝑒𝑎 = (. 0127𝑚) ∗ (. 0127𝑚) = .00016129 𝑚2 𝐹 = 11112.881 𝑁
11112.881 𝑁 𝑁
𝐺𝐹𝑇 = (𝐹 ⁄𝑉𝑅𝐼𝑂 ) = = 32579.54
0.3411 𝑉 𝑉

9. The myRIO A-side can only measure from 0 to 5 V. To avoid outputting negative voltages from the
transducer circuit in case of drift, the potentiometer is used to balance the Wheatstone bridge so that
the offset voltage is 0.1 V. What is the calibration equation for the force transducer you would enter
into the VI where the voltage read by the myRIO is 𝑉𝑅𝐼𝑂 and the offset voltage is 0.1 V?
𝐹 = 𝐺𝐹𝑇 ∗ (𝑉𝑅𝐼𝑂 + 0.1)

Force Transducer Calibration

During the lab you will be checking the theoretical design calibration equation by characterizing the individual
components of the force transducer and signal path. There may be additional effects from the rig itself that
could influence your calibration and the precision of your results. You will check this by attaching a crane scale
to your rig and using it is a measurement standard for calibration. A sample set of calibration data using the
crane scale can be found in “calibration_data.mat” where “V_rio” is the measured myRIO voltage from the force
transducer and “m_scale” is the mass displayed by the scale in kilograms.

10. What is the gain of the force transducer in N/V?

Using the xFit command in Matlab,

Using matlab xFit Coefficients (with 95% confidence bounds):

p1 = 2.101e+04
Gain=2.101e+04 N/V

11. What is the precision of the force transducer (K = 2 prediction interval) in N? (Hint: Use the rmse value
from the fit() function “goodness” for your calculation, not the confint() function. The answer should be
in the form of ±X N.)
rmse: 140.9800 = SSD
SSDmean=49.8439
2.101e4 (+ and -)2*SSD
2.101e4 (+ and -) 281.9600 Newtons

5
ME 3057 Homework 6 Force Transducer

12. Plot the calibration data, calibration curve, and prediction interval with the measured voltage on the x-
axis and the force on the y-axis.
Plot below

Figure 3. Gain of Experimental Data

Failure Testing
Variation in manufacturing and materials can result in variations in the properties of final products.
Manufactures typically conduct destructive testing to determine if their products meet the design goals or
determine the rated limits. This week you will conduct destructive testing on a climbing sling and using the
combined results from your lab section determine if the sling meets its designed rating. A sample set of sling
failure testing data can be found in “sling_testing_data.mat”, where “time_s”, “displacement_mm”, and
“force_N” are data for a single sling failure test and “ultimate_kN” are the ultimate strengths for additional test.
6
ME 3057 Homework 6 Force Transducer

13. During testing there will initially be a small amount of slack in the sling before it is loaded. Plot the tensile
force versus the change in length of the climbing sling. (Hint: Segment the data to only show from when
the sling is initially loaded to failure as there is excess data before and after.)
Plot below

Figure 1: Plot showing Force Vs. length increase

14. What is the ultimate strength of the climbing sling?

Ultimate Strength= 31,260 Pa shown in Figure 1 on the Y axis.

7
ME 3057 Homework 6 Force Transducer
15. Using your result from question 14 and the previous test results, what is the mean ultimate strength of
the climbing sling? What is the standard deviation? (Hint: Do not use the fit() function for this as there
is only one degree of freedom. Try using the mean() and std() functions.)
Mean ultimate strength=30,487 Pa STD=378.38 Pa

16. If the measurements were recorded using the force transducer from questions 10 to 12, what is K = 2
95.4% prediction interval for the sling ultimate strength? (Hint: Use the square root sum of squares to
account for the uncertainty in the force transducer measurements and the variation in the slings.)
prediction_interval= mean_Ultimate_kN +and - total_uncertainty_K_2 %kN
How do you type + - with 1 over the other without a division sign?
Total uncertainty is both the mean absolute value distance from each ultimate strength data point
from mean and the standard deviation of the ultimate strength. Square both then add them together.
You then square root the answer for the total uncertainty. Total_uncertainty = 5.15 kN
Precision = mean_ultimate_kN

Prediction interval = 31.339kN +and- 5.15 kN

= 26.1890kN
= 36.4890kN

17. If the slings are rated for a maximum load of 17 kN, what is the factor of safety in the sling design?

Factor of safety= mean_Ultimate_Strength_kN / sling_load_rating = 1.8435

𝜎𝑢𝑙𝑡𝑖𝑚𝑎𝑡𝑒
𝑛=
𝜎𝑅𝑎𝑡𝑒𝑑

%Repp Ledger
%10/14/18
%HW6Q10 ME-3057 G
%10. What is the gain of the force transducer in N/V?
load('calibration_data.mat')
plot(V_rio,m_scale,'o')
gravity=9.81 %m/s^2
Force=m_scale.*gravity
[xFit,stats] = fit(V_rio,Force,'poly1') %polyfit 1st order curve (line) to data
xlabel('Measured Voltage (V)')
ylabel('Force(N)')
title(['T'])
gain=(1.462e4-0)/(0.8-0.1) %manual to compare
SSD=stats.rmse

% Question 12 (uses data above here)

xFitPIMean = predint(xFit,V_rio,0.954,'functional','off');

8
ME 3057 Homework 6 Force Transducer
% Matlab built in prediction interval calculator(.954 gives us the approximate 2*s interval)
xFitPIObsv = predint(xFit,V_rio,0.954,'observation','off');
% (.954 gives us the approximate 2*s interval)
figure(2)
plot(xFit,V_rio,Force,'o'); % plot discrete data points
hold on
plot(V_rio,xFitPIMean,'--k')%Mean PI upper and lower bound
plot(V_rio,xFitPIObsv,'--g')%Observation PI upper and lower bound
hold off
grid on
xlabel('Measured Voltage [V]')
ylabel('Force [N]')
title(['Prediction Intervals for Measured Voltage Vs Force'])
legend('Measured Data','Expected Value','Mean PI Upper Bound',...
'Mean PI Lower Bound','Observation PI Upper Bound',...
'Observation PI Lower Bound','Location','NorthWest')
set(gca,'FontSize',14); % bigger fonts

9
ME 3057 Homework 6 Force Transducer

%Repp Ledger
%10/15/18
%HW6Q13 ME-3057 G
clc,clear
load('sling_testing_data.mat')
displacement_m=(displacement_mm(30315:50339).*1/1000)';
force_N=force_N(30315:50339);
ultimate_kN=ultimate_kN(:,1);
plot(displacement_m,force_N);
xlabel('Displacement (m)')
ylabel('Force(N)')
title(['Tensile Force Vs. Change in Length'])
grid on
figure (2)
displacement_m=(16916:20025);
displacement_m=displacement_m';
Ultimate_strength=max(force_N)
Force_segment=force_N(16916:20025);
plot(displacement_m,Force_segment);

10
ME 3057 Homework 6 Force Transducer

plot(displacement_m,Force_segment);
hold on
[xFit,stats] = fit(displacement_m,Force_segment,'poly1')
xlabel('Measured Voltage (V)')
ylabel('Force(N)')
title(['Voltage Vs. Force'])
% Question 12 (uses data above here)
xFitPIMean = predint(xFit,displacement_m,0.954,'functional','off');
% Matlab built in prediction interval calculator(.954 gives us the approximate 2*s interval)
xFitPIObsv = predint(xFit,displacement_m,0.954,'observation','off');
% (.954 gives us the approximate 2*s interval)
figure(3)
plot(xFit,displacement_m,Force_segment,'o'); % plot discrete data points
hold on
plot(displacement_m,xFitPIMean,'--k')%Mean PI upper and lower bound
plot(displacement_m,xFitPIObsv,'--g')%Observation PI upper and lower bound
hold off
grid on
xlabel('Measured Voltage [V]')
ylabel('Force [N]')
title(['Prediction Intervals for Measured Voltage Vs Force'])
legend('Measured Data','Expected Value','Mean PI Upper Bound',...
'Mean PI Lower Bound','Observation PI Upper Bound',...
'Observation PI Lower Bound','Location','NorthWest')
set(gca,'FontSize',14); % bigger fonts

11
ME 3057 Homework 6 Force Transducer

12
ME 3057 Homework 6 Force Transducer

Published with MATLAB® R2018a

%Repp Ledger
%10/15/18
%HW6Q16 ME-3057 G
%QUestion 16. If the measurements were recorded using the force transducer
%from questions 10 to 12, what is K = 2 95.4% prediction interval for the sling ultimate
%strength? (Hint: Use the square root sum of squares to account for the uncertainty in the
%force transducer measurements and the variation in the slings.)
clc,clear
load('sling_testing_data.mat')
mean_Ultimate_kN=mean(ultimate_kN);
U1=abs(mean_Ultimate_kN-ultimate_kN);
U1=mean(U1);
SDD_ult_str=(std(ultimate_kN)) %kN
total_uncertainty_K_2=sqrt((U1)^2+ (2*SDD_ult_str)^2)
prediction_interval= mean_Ultimate_kN + total_uncertainty_K_2 %kN
prediction_interval= mean_Ultimate_kN - total_uncertainty_K_2 %kN
sling_load_rating=17 %kN
factor_of_safety=mean_Ultimate_kN/sling_load_rating

13
ME 3057 Homework 6 Force Transducer

Published with MATLAB® R2018a

14