ME444: Analysis and Design of Mechanisms
Lab 5
(Vibrations of Multi-Degree of Freedom Systems)
Group 8
Date: April 12, 2025
Sr. No Name Roll Number
1 Ayush Lonakadi 23B2253
2 Varun Inamdar 23B2203
3 Aagam Kasaliwal 23B2215
4 Lakshaditya Singh 23B2230
5 Anirudh Maindargikar 23B2219
1.Introduction:
● Objectives
The main goal of this lab was to understand how mechanical systems with
one or more degrees of freedom behave under free vibrations. We aimed to:
● Measure the natural frequency of a simple mass-spring system
(1-DoF).
● Estimate spring stiffness using both dynamic and static approaches.
● Use FFT to analyze the vibration signal and extract frequency data.
● Extend this understanding to a two-mass, two-spring system (2-DoF)
and observe the coupled behavior.
● Compare experimental results with theoretically predicted natural
frequencies using matrix methods and eigenvalue analysis.
● Background of Experiment
Many physical systems in engineering exhibit vibratory motion, especially
when displaced from their equilibrium position and released. These systems
are often modeled as second-order differential systems due to the
involvement of mass (inertia) and stiffness (restoring force), as seen in
typical mass-spring systems.
This lab offers hands-on exposure to the principles governing such
oscillatory systems. Students will develop an intuitive and analytical
understanding of fundamental vibration mechanics by exploring single and
coupled mass-spring systems. The experiment also emphasizes measurement
techniques, data accuracy, and the transition from time-domain analysis to
frequency-domain analysis via FFT.
● Terminologies
Here are some of the key terms that were central to this lab:
● 1-DoF / 2-DoF: Degree of freedom refers to the number of
independent ways a system can move. A 1-DoF system moves in one
direction; a 2-DoF system has two independent motions.
● Natural Frequency: The frequency at which a system vibrates freely
when disturbed and left alone.
● Spring Stiffness (k): Indicates how much force a spring exerts per
unit displacement.
● Free Vibration: Vibration occurs without any ongoing external force
due to the system's initial disturbance.
● FFT (Fast Fourier Transform): A tool that breaks down a
time-domain signal into its frequency components — very useful in
vibration analysis.
● Eigenvalues: Values obtained from solving a matrix equation give us
the system’s natural frequencies in multi-degree-of-freedom systems.
2. Methodology:
● 1-DoF Free Vibration:
○ Suspended a known mass from a spring attached to a fixed
bracket.
○ Displaced the mass slightly and released it.
○ Timed >10 complete oscillations using a stopwatch; repeated at
least 3 times.
○ Calculated the average natural frequency using total time and
number of cycles.
● Dynamic Stiffness Calculation:
○ Used measured frequency and known mass to compute
stiffness:
● Static Stiffness Calculation:
○ Measured spring extension y due to mass.
○ Calculated stiffness using:
● FFT-Based Frequency Measurement:
○ Mounted an accelerometer on the mass.
○ Recorded vibration data via LabJack and performed FFT.
○ Extracted natural frequency and recalculated stiffness.
● 2-DoF Free Vibration:
○ Connected two masses and two springs in series.
○ Displaced both masses together (in the same and opposite
directions).
○ Timed oscillations to estimate two natural frequencies.
● Theoretical Analysis:
○ Derived equations of motion and formed matrix system.
○ Solved for eigenvalues to find theoretical natural frequencies.
○ Compare theoretical results with experimental values.
3. Results
1. Spring 1
a. Free response of a 1-DOF system
● m = 0.310 kg
Table 1: Results for free response of 1-DOf system for spring 1
Sr.no Number of Time taken (s) Time period (s)
oscillations
1 15 7.65 0.51
2 15 7.62 0.51
3 15 7.74 0.52
● Time period of oscillation = 0.51s
● Using the relation, k = (2πf)2.m, we have: k = 47.00N/m
b. Static displacement method
● m = 0.310kg
● g = 9.81 m/s2
● Natural length of the spring = 16.0 cm
Table 2: Results of static displacement method for Spring 1
Sr. no Mass (kg) Final length (cm) Extension (cm)
1 0.310 22.4 6.4
2 0.310 22.4 6.4
3 0.310 22.5 6.5
● Extension = 6.4 cm
● By the relation, k = mg/y, we have: k = 47.52 N/m
2. Spring 2
a. Free response of a 1-DOF system
● m = 0.316 kg
Table 3: Results for free response of 1-DOF system for spring 2
Sr.no Number of Time taken (s) Time period (s)
oscillations
1 15 6.44 0.43
2 15 6.45 0.43
3 15 6.42 0.42
● Time period of oscillation = 0.43s
● Using the relation, k = (2πf)2.m, we have: k = 70.65N/m
c. Static displacement method
● m = 0.316kg
● g = 9.81 m/s2
● Natural length of the spring = 16.0 cm
Table 4: Results of static displacement method for Spring 2
Sr. no Mass (kg) Final length (cm) Extension (cm)
1 0.316 20.4 4.4
2 0.316 20.4 4.4
3 0.316 20.5 4.5
● Extension = 4.4 cm
● By the relation, k = mg/y, we have: k = 70.45 N/m
d. FFT on accelerometer output for spring 2
Figure 1: FFT on accelerometer output for spring 2 (First iteration)
Figure 2: FFT on accelerometer output for spring 2 (Second iteration)
Figure 3: FFT on accelerometer output for spring 2 (Third iteration)
Table 5: FFT data obtained for Spring 2
Sr. no Mass (kg) Frequency (Hz)
1 0.316 2.375
2 0.316 2.333
3 0.316 2.400
● Frequency after FFT: 2.370 Hz
● Using the relation, k = (2πf)2.m, we have: k = 70.00 N/m
3. Free response of a 2-DoF translational system
a. Displacement of both masses in the same direction
Figure 4: Schematic representation of the setup
Table 6: Time period for in-phase oscillations of the 2-DoF system
Sr. no Number of oscillations Time taken (s) Time period (s)
1 10 7.83 0.783
2 10 7.85 0.785
3 10 7.85 0.785
b. Displacement of both masses in the opposite direction
Table 7: Time period for out-of-phase oscillations of the 2-DoF system
Sr. no Number of oscillations Time taken (s) Time period (s)
1 10 3.72 0.372
2 10 3.90 0.390
3 10 3.45 0.345
4. Theoretical Natural frequencies
Figure 5: Free-body diagram for the situation
From Newton’s second law, we have
In the matrix form, we have
Putting y1 = Y1cos(ωt) and y2 = Y2cos(ωt), we have
Since the amplitudes cannot be zero, the determinant of the matrix
must be zero:
From the previous results, we have
● k1 = 47.00 N/m
● k2 = 70.00 N/m
● m1 = 0.310 kg
● m2 = 0.316 kg
On solving the quadratic equation for ω2, we have ω2 = 62.622 or 535.322
Taking ω2 = 62.22, we get ω = 7.89 rad/s for in-phase oscillations
● T = 2π/ω = 0.796s
Taking ω2 = 535.522, we get ω = 23.137 rad/s for out-of-phase oscillations
● T = 2π/ω = 0.271s
4. Discussion
This section critically examines the experimental results obtained for the
1-DoF and 2-DoF mass-spring systems and compares them with theoretical
predictions. Any discrepancies are analyzed, and the accuracy and
consistency of the different measurement methods are discussed.
Spring Stiffness from 1-DoF Free Response
For Spring 1, the time period obtained from the free oscillation method was
consistently 0.51 s. Using this, the calculated dynamic stiffness was 47.00
N/m. This closely matched the extension method's static stiffness value of
47.52 N/m. The negligible deviation (~1.1%) suggests that both methods are
accurate and mutually validating. The minor difference can be attributed to
human error during stopwatch timing or measurement of spring extension.
Similarly, Spring 2's time period was approximately 0.43 s, leading to a
dynamic stiffness of 70.65 N/m. The static method gave a comparable value
of 70.45 N/m. The error (~0.3%) is smaller here, indicating highly consistent
and reliable measurements across dynamic and static approaches.
These results support the assumption of ideal spring behavior (Hooke’s law)
and confirm that our mass-spring systems operated linearly.
FFT-Based Natural Frequency Identification
For Spring 2, FFT analysis of the accelerometer data yielded an average
frequency of 2.370 Hz across three trials. This corresponds to a stiffness of
70.00 N/m using the same mass (0.316 kg), which is consistent with both the
static (70.45 N/m) and free vibration methods (70.65 N/m). Despite relying
on electronic sensors and signal processing, the FFT-based method showed
excellent agreement with manual methods. This demonstrates the
effectiveness of FFT for accurate frequency analysis, especially when
high-precision timing tools are unavailable.
The small deviation (~0.9%) among methods is likely due to the resolution
of the FFT output and potential noise in the accelerometer data. However,
the consistency validates the practical reliability of frequency-domain
analysis for mechanical vibrations.
Natural Frequencies of the 2-DoF System
For the coupled 2-DoF system, two types of motions were studied: in-phase
and out-of-phase oscillations.
● In-Phase Motion: The measured average time period was 0.784 s.
Theoretically, solving the eigenvalue problem gave a time period of
0.796 s. The small discrepancy (~1.5%) is acceptable and can be
attributed to simplified assumptions in the analytical model, such as
ignoring damping and assuming perfectly rigid connections.
● Out-of-Phase Motion: The experimental average time period for
out-of-phase oscillations was 0.369 s, while the theoretical value
calculated from eigenvalue analysis was 0.271 s. This discrepancy
(~36%) is more significant and may stem from the difficulty in
exciting a purely out-of-phase mode manually. Mixed-mode behavior
or non-negligible damping effects not accounted for in the theoretical
model could also contribute to the mismatch. Furthermore, the higher
frequency of this mode may have exaggerated any timing errors,
especially if manual measurements were used.
Some mismatches might also arise due to difficulty in manually generating
pure mode shapes. In a laboratory setting, achieving exact in-phase or
out-of-phase excitation is challenging, often resulting in mixed-mode
behavior and slight shifts in observed periods.
General Observations and Unexpected Results
While most results aligned well with theoretical predictions, the larger
deviation observed in the out-of-phase 2-DoF mode suggests that precision
in experimental setup becomes more critical at higher frequencies or for
more complex modes. Future iterations could improve mode excitation
precision and incorporate damping considerations in analysis.
However, a few observations merit discussion:
● Due to measurement uncertainties, minor discrepancies between static
and dynamic stiffness values are expected.
● Manual timing using a stopwatch introduces human reaction error;
digital data acquisition (e.g., accelerometers with FFT) offers more
precision and reproducibility.
● In the 2-DoF system, ensuring purely decoupled modes of vibration
remains a challenge and could be improved in future setups using
controlled initial displacements and damping consideration.
5. Conclusions:
1. High Agreement Between Static and Dynamic Stiffness Measurements:
For both springs, the stiffness values obtained through static displacement
and dynamic free vibration methods were in close agreement (within ~1%).
This confirms the reliability of both approaches in estimating spring
constants and validates the use of linear spring-mass models under small
displacements.
2. FFT-Based Frequency Analysis is Highly Reliable:
The natural frequency obtained through FFT analysis of accelerometer data
for Spring 2 (2.370 Hz) yielded a stiffness value of 70.00 N/m, which
closely matched results from manual timing and static extension methods.
This reinforces the suitability of FFT for vibration characterization,
especially when higher accuracy is required.
3. Good Agreement for In-Phase 2-DoF Mode, but Discrepancy for
Out-of-Phase:
The in-phase oscillation of the 2-DoF system showed excellent agreement
with the theoretical time period (0.784 s experimental vs. 0.796 s
theoretical). However, the out-of-phase mode had a notable discrepancy
(0.369 s experimental vs. 0.271 s theoretical), likely due to imperfect mode
excitation and unmodeled damping effects.
4. Validation of Analytical Methods via Experimental Results:
The matrix formulation and eigenvalue solution for natural frequencies of
multi-degree-of-freedom systems provided accurate predictions. This
demonstrates the practical applicability of theoretical tools taught in class for
real-world system analysis.
5. Manual Timing is Effective but Limited in Precision:
While stopwatch-based measurements proved sufficient for educational and
approximate analysis, their accuracy is inherently limited compared to
sensor-based approaches.
Limitations and Suggested Improvements
Limitations:
● Inaccurate Excitation of Out-of-Phase Motion: Manual initiation of
perfectly out-of-phase motion is challenging, which likely caused the
observed mismatch with theory.
● Human Error in Manual Timing: Stopwatch-based measurements are
subject to human reaction time, which introduces error, especially over
short-duration oscillations.
● Inexact Mode Shape Excitation in 2-DoF System: Achieving pure
in-phase or out-of-phase displacement by hand is difficult, which may lead
to mixed-mode vibrations and slight deviations in observed periods.
● Sensor Noise and FFT Resolution: Although FFT analysis provides greater
accuracy, the accelerometer data may still carry noise, and the sampling rate
and window size limit the frequency resolution.
Suggested Improvements:
● Automate Timing Using Optical or Motion Sensors: Replacing manual
stopwatch measurements with photogates or displacement sensors would
improve time period precision.
● Use Controlled Initial Displacement Mechanisms: Implementing guided
setups for in-phase and out-of-phase excitation can help isolate pure modes
in the 2-DoF experiment.
● Higher-Resolution Data Acquisition for FFT: Increasing the sampling rate
and using windowing techniques can enhance the clarity and accuracy of the
FFT spectrum.
● Incorporate Damping Effects in Analysis: Extending the theoretical model
to include damping would make it more representative of real systems and
improve prediction accuracy
6. Contributions:
Name Roll Number Contributions
Varun Inamdar 23B2203 Introduction, Data colllection
Aagam Kasaliwal 23B2215 Results, plots
Ayush Lonakadi 23B2253 Discussion, data collection
Lakshaditya Singh 23B2230 MATLAB scripting
Anirudh Maindargikar 23B2219 Results, Conclusion
7. References:
● https://moodle.iitb.ac.in/mod/resource/view.php?id=59395 (Moodle -
Lab 5 - 12th April 2025)
● https://en.wikipedia.org/wiki/Fast_Fourier_transform#:~:text=A%20f
ast%20Fourier%20transform%20(FFT,frequency%20domain%20and
%20vice%20versa. (FFT - Wikipedia - 12th April, 2025)
