Vehicle Ride Comfort Testing: Noise and Vibration Analysis of a Mini Cooper R56

This study investigates the ride comfort of a Mini Cooper R56 using synchronized interior-noise and whole-body vibration measurements based on ISO 5128:2023 and ISO 2631-1:1997. Using Dewesoft measurement hardware and software, the tests evaluate cabin noise, vibration exposure, ride comfort, and motion-sickness indicators under stationary and real-world driving conditions. The results are also compared with relevant European exposure limits to assess both acoustic and vibration performance.

Introduction

My study aims to measure and analyze a vehicle’s ride comfort (Mini Cooper R56), following: 

• ISO 5128:2023, “Acoustics — Measurement of noise inside motor vehicle,” and

• ISO 2631-1:1997, “Mechanical vibration and shock — Evaluation of human exposure to whole-body vibration — Part 1: General requirements.”

Subsequently, I will compare the results with the two European Directives concerning the limit exposure values, both for acoustics and vibration:

• Directive 2003/10/EC: Minimum health and safety requirements regarding the exposure of workers to the risks arising from physical agents (noise).

• Directive 2002/44/EC: Minimum health and safety requirements regarding the exposure of workers to the risks arising from physical agents (vibration)

I developed this project as a thesis for the Bachelor’s degree in Vehicle Engineering at the University of Modena and Reggio Emilia (UNIMORE).

Meeting ISO standards for ride comfort

To perform measurements and analyses in accordance with ISO standards, it was necessary to comply with the two relevant normative documents summarized below.

ISO 5128:2023

The microphone position, which you must calibrate at the beginning of each measurement, should be near the headrest. 

You have to perform the tests under very strict weather conditions: no rain or wind (because these conditions could affect the test) and within the range of -10°C to 40°C.

There are several types of noise measurement tests, both dynamic and static. In our case of interest, the Standstill measurement required the vehicle to be at least 20 meters from every building to avoid influencing the results.

The Standstill Condition requires three different acoustic level measurements:

• Background Noise, at the beginning and end of the test (engine off);

• Engine on, idle, with the AC at minimum power

• Engine on, idle, with the AC at maximum power

Then, to finalize the results, we used the latter two values to determine the sound equivalent level in accordance with ISO.

As we shall see in the latter chapters, we decided not to measure the AC at minimum power. Instead, we measured the sound level with the AC OFF to enhance the difference between the phases and, therefore, the contribution of the AC.

ISO 2631-1:1997

The sensor must be positioned at the point of interaction between the worker and the machine: in our case, at the center of the driver’s seat.

Three different analysis methods are defined:

1. Basic Evaluation (RMS Method): The weighted root-mean-square acceleration, denoted as, aRMS is calculated. This value represents the average acceleration over the entire measurement duration and is often considered sufficient to describe human exposure to vibration.

2. Running RMS Method (MTVV): This method calculates a short-duration running-time window with a 1-second integration time constant. The Maximum value obtained over the entire measurement interval is the reference value of the Maximum Transient Vibration Value (MTVV). This method is extremely useful when the signal contains shocks.

3. Fourth Power Vibration Method (VDV): The calculation is based on the fourth power of the instantaneous acceleration, producing the Vibration Dose Value (VDV). This value is significantly more sensitive to the RMS acceleration.

The choice of the most appropriate method depends on the nature of the signal, specifically the presence of shocks or impulsive peaks, which you determine by calculating the Crest Factor (CF): if CF>9, the basic Evaluation method is inefficient to assess the effectiveness of the results.

However, the most widely used method is the Fourth Power Vibration Method because it is included in the European directive.

Another important aspect defined in this standard concerns motion sickness: although it is difficult to compare results objectively because everyone reacts differently, you calculate the Motion Sickness Dose Value (MSDV), which accounts for vertical acceleration.

However, as defined in several recent scientific papers, the primary cause of motion sickness while driving is lateral acceleration.

Equipment used

We utilized the following professional instrumentation to ensure synchronized high-resolution data acquisition:

• DAQ System: Dewesoft SIRIUS 16xACC, a modular data acquisition system with high-end signal-conditioning amplifiers; see Figure 3.

• Acoustic Sensor: GRAS 146AE Free-field Rugged Microphone (IEPE/TEDS);

• Vibration Sensor: PCB 356A03 Triaxial ICP® Accelerometer;

• Inertial Platform: Dewesoft Navion i2 (GPS/IMU) for route mapping and speed monitoring;

• Software: DewesoftX with Sound Level Meter and Human Body Vibration modules.

The experimental setup followed the strict ISO guidelines:

• Acoustics (ISO 5128): The microphone was positioned at the level of the driver's headrest, in the seat's plane of symmetry, to capture interior noise during standstill tests (see Figure 2).

• Vibrations (ISO 2631-1): The triaxial accelerometer was secured at the center of the seat interface, capturing longitudinal (x), lateral (y), and vertical (z) accelerations transmitted to the driver, see Figure 2.

In addition, we mounted the Inertial Measurement Unit (IMU) on the vehicle's roof (see Figure 1).

DewesoftX modules, frequency analysis, and basic statistic tools

List of relevant modules and mathematical tools that facilitated both the measurement procedures and the subsequent analysis:

• Sound Level Meter: Used for instantaneous acquisition and scaling of sound pressure to dB(A), enabling calculation of the equivalent sound level perceived by the human ear, and supporting various time weightings. The output parameters chosen are the instantaneous dB(A) level and the equivalent dB(A) level, which calculates the mean across the whole measurement  LEQ,DeweX . We will compare this value with the equivalent sound level calculated from the equation defined in ISO 5128:2023 for the Standstill condition (see Figure 5).

• Human Body Vibration (HBV): A specialized module for whole-body vibration analysis in accordance with ISO 2631-1. It automatically handles frequency weighting filters and the specific multiplication factors for different axes. Not only can you select the parameters defined in ISO 2631-1, but this module also includes other parameters crucial for lumbar spine assessment, defined in ISO 2631-5 (see Figure 6). 

• Navion i2 Tool: Integrated for processing GPS and inertial data, allowing for precise route mapping and synchronized monitoring of speed and position;

• Basic Statistics: used for two key functions, calculating the Crest Factor (to determine the necessity of the VDV method over the standard RMS) and determining the Average Speed of the vehicle during test runs (to demonstrate compliance with Italian street legislation).

• FFT (Fast Fourier Transform): An efficient algorithm used for both acoustic and vibration signals to identify dominant frequencies. In the Acoustic case, we used it twice: in the Overall condition to determine the main frequencies over the entire measurement interval, and in the time-interval condition to determine the difference between the AC-ON and AC-OFF cases. Meanwhile, in the vibration analysis, we only calculated the Overall FFT. The three of them used Hanning’s windowing and a 66.67% overlap to avoid information loss.

• Octave Analysis / CPB (Constant Percentage Bandwidth): Applied specifically to acoustic data using 1/3-octave filters to obtain a spectral representation that closely aligns with human auditory perception.

The test structures in compliance with the ISO standards

We conducted the experimental tests on February 20, 2026, in the district of Vitinia, located within the municipality of Rome. Dividing the testing procedures into two distinct categories: a stationary acoustic analysis and a dynamic vibration analysis, we performed both using the professional data acquisition hardware and software previously described.

Acoustic test structure

We structured the acoustic measurement as a static standstill test conducted in an outdoor parking lot. To prevent significant sound reflections, we selected a site that kept the vehicle at least 20 meters from large buildings or structures. The test strictly followed the ISO 5128 standard and was organized into the following procedural sequence:

1. Environmental conditions

We closed all windows and sunroofs to isolate the cabin, and initially shut off the engine. The microphone was at the headrest level.

2. Initial background noise

We made a 10-second recording with the engine off to establish the Background noise.

3. Engine Idle (AC OFF)

We started the engine and maintained its minimum idle speed for 30 seconds with the air conditioning system off.

4. Engine Idle (AC MAX)

We set the air conditioning to maximum power for 30 seconds while the engine continued to idle

5. Final Background Noise

We turned off the engine for a final 10-second background noise recording to ensure the environment remained stable.

Vibration test structure

We conducted the vibration analysis in accordance with the ISO 2631-1:1997 standard for whole-body vibration. 

We carried out the measurements over a mixed urban circuit of approximately 1.3 km in Vitinia, see Figure 9. We selected this route for its diverse technical features, including a traffic light, an uphill section, a downhill section, a straight path, and, in particular, irregular road surfaces designed to transfer the most high-impact shocks to the driver.

The structure of the test was as follows:

1. Stationary Phase: Each test run began with a 20-second measurement at engine idle while the vehicle was stationary.

2. Dynamic Phase: The vehicle traveled the 1.3 km mixed route at an average speed of 20 km/h, adhering to local traffic regulations.

3. Final Phase: Each run ended with another 20-second stationary idle measurement.

Analysis and results

Acoustic analysis

Table 1 reports all the scalar quantities of interest obtained from the measurements, all with A-type weighting.

You can observe that turning on the air conditioning leads to an approximately 14 dB increase in A-weighted sound pressure. 

Another evident difference we can see is between the values of the equivalent sound level calculated by the software LEQ,DeweX and those calculated in accordance with the ISO 5128:2023 standard LStandstill,ISO. We found the cause for this difference in the calculation criteria:

• LEQ,DeweX calculates a mean, given the samples, over the whole test;

• LStandstill,ISO is characterized by a complex dependence on the two detected values, LAC,OFF and LAC,ON,  (the maximum value at each phase), where the first one has a multiplicative factor equal to 9 times that assigned to the first, as seen in the formula.

Ultimately, with reference to European Directive 2003/10/CE regarding noise exposure in workplaces, it can be concluded that, in the case of Standstill analysis according to the ISO 5128:2023 standard, the measured vehicle is well within the exposure limits: in fact, the average value obtained, whether it the equivalent ISO Level, LStandstill,ISO ,  or the one calculated by DewesoftX, LEQ,DeweX , is much lower than the exposure limit value, equal to 87 dB(A) defined by the directive.

The graph in Figure 10 represents the Overall FFT (over the entire measurement interval) in the Hz range of interest. However, for the FFTs, the dB(Z) value is plotted on the ordinate, not dB(A), which corresponds to the unweighted sound level detected by the microphone. 

In this range, excluding the 0 Hz peak at 100 dB, which is probably an artifact introduced by the band, a peak of 87.202 dB is observed at the characteristic frequency of 24.41 Hz. This value, also considering the resolution, coincides with the second mechanical combustion order; therefore, it is the dominant combustion frequency. Thanks to this result, it is possible to trace back to the (average) idle speed of the engine during the test: 

The second mechanical order represents an event that repeats exactly twice per crankshaft rotation. In this vehicle, equipped with a 4-stroke, 4-cylinder inline engine, the thermodynamic cycle is completed every 2 shaft rotations; therefore, there are 2 combustion cycles per shaft rotation.

Proceeding, a comparison is made between the two test moments, with air conditioning on and off, in Figures 11 and 12. The difference between the two 24.41 Hz peaks is only 4 dB (linear, unweighted). What actually contributes to the 14 dB(A) difference seen in Table 1 is the presence of broadband noise (additive, constant noise across all frequencies, similar to white noise but of a fluid-dynamic nature) in the measurement with the air conditioning on. 

Therefore, due to broadband noise, the dB(A) value detected in the second case (Figure 12) differs appreciably from that in the first case (Figure 11) because of contributions at higher frequencies, which increase the dB(A) equivalent value. 

As a final consideration, Figure 13 shows the octave analysis of the sound level detected in the Overall CPB over the Hz interval. Analyzing the detected peak value, we can conclude that this magnitude does not differ significantly from the analysis carried out, again in the Overall, using the FFT in Figure 10. The slight difference in frequency and sound-level values is due to the lower frequency resolution of the FFT and the higher bandwidth; therefore, acoustic energy absorption is higher at 25 Hz (the nominal frequency for ISO standards) than at 24.41 Hz.

Vibration analysis

A note on the methodology: The accelerometer I used has a manufacturer-defined measurement interval with a lower limit (2 Hz) that exceeds the frequency range of interest for motion sickness ([0.1, 0.5] Hz), potentially introducing artifacts up to about 2 Hz. Despite this hardware limitation, we performed the measurements in strict accordance with the ISO 2631-1:1997 standard and still implemented the MSDV calculation to complete the analysis chain.

Considering the average values from Table 2, ride comfort and the exposure value are evaluated according to the ISO 2631-1:1997 standard and the European Directive 2002/44/CE, respectively:

• aRMS: vehicle (and road surface) generates average accelerations (vibrations) classified as Fairly Uncomfortable;

• VDV: below the action value (and exposure limit);

• A(8): below the action value (and exposure limit).

Due to crest factors greater than 9 detected in the mixed route, we applied the fourth-power vibration method, rather than the MTVV method, to enable comparison of the obtained value with the limit values specified in the European directive. 

The large difference between the average acceleration value, MTVV, and VDV, highlights the insensitivity of the first to the severe impulsive peaks present in the detection, due to impacts or bumps: the ineffectiveness of the standard evaluation method in these conditions and the need for more complex and sensitive tools, as also required by the reference standards, is thus confirmed.

Figure 14 represents the Overall FFT analysis of the accelerations detected in all three directions. The area of greatest interest that emerges from the graph is defined by the peak, common to the three directions, at 2.67 Hz, which therefore represents the natural frequency of the vehicle's suspended masses. This value, significantly higher than the ideal of about 1.5 Hz, highlights the high rigidity of the vehicle's setup, which penalizes ride comfort in favor of better driving dynamics.

Figure 15 shows only the field of interest for measuring motion sickness and the acceleration along the vertical z-axis. As noted in the methodological note, the MSDV values calculated by the software are inconsistent. No energy peak characterizes the noise present in the interval; rather, a slight slope variation at 0.38 Hz is observed, which could be due to the instrumental background noise of the accelerometer, which, as already mentioned in the note, has a cut-off frequency starting from 2 Hz.

Although the measurement and analysis of vibrations in the cabin strictly follow the ISO 2631-1:1997 standard, a major issue emerges from the most recent publications: motion sickness is mainly influenced by low-frequency lateral accelerations of the vehicle, not by vertical ones, as defined by the standard. This observation highlights an underestimation in the ISO standard, still in force, regarding the calculation of the MSDV in the automotive field, or more generally in land transport on rubber. 

Therefore, for completeness (and despite the methodological note), Figure 16 shows the FFT of the MSDV range of interest, now comparing the lateral y- and vertical z-accelerations. The lateral acceleration is higher than that on the z-axis; therefore, the "updated" MSDV could be higher, given an appropriate weighting factor.

Conclusions

The project successfully demonstrated that the Dewesoft ecosystem provides a seamless solution for complex NVH challenges. By integrating SIRIUS hardware with the Navion i2 inertial platform and DewesoftX software, we synchronized high-speed acoustic and vibration data with GPS positioning data in a single workspace. The integrated ISO-compliant math modules eliminated the need for fragmented post-processing, transforming a rigorous academic study into an efficient engineering workflow.

Acknowledgements

My deepest thanks go to Elvira Rufolo, Emanuele Burgognoni, Davide Carniani, Riccardo Petrei, and the entire Dewesoft Italy team, whose expert support and unfailing patience were instrumental in shaping and executing the experimental procedures in this work. Quite simply, this study would not exist without their contribution.

Sources and references

International Standards

• ISO 2631-1:1997. Mechanical vibration and shock — Evaluation of human exposure to whole-body vibration — Part 1: General requirements. International Organization for Standardization, Geneva, Switzerland.

• ISO 5128:2023. Acoustics, Measurement of interior vehicle noise—International Organization for Standardization, Geneva, Switzerland.

European Union Directives

• Directive 2002/44/EC. Minimum health and safety requirements regarding workers' exposure to the risks arising from physical agents (vibration). Official Journal of the European Communities.

• Directive 2003/10/EC. Minimum health and safety requirements regarding workers' exposure to the risks arising from physical agents (noise). Official Journal of the European Union.

Scientific Literature, regarding Motion Sickness

• Dacova, D. (2021). Ride comfort in road vehicles: a literature review. International Scientific Journal "Trans and Motauto World", 6, 60-63.

Dewesoft Hardware & Software Resources

• Dewesoft d.o.o. SIRIUS Data Acquisition System (DAQ).  

• Dewesoft d.o.o. (2024). Introducing Navion i2: The Ultimate Automotive INS.  

• Dewesoft Academy. Sound Level Measurement. 

• Dewesoft Academy. Human Body Vibration.

• Dewesoft Academy. FFT Spectrum Analysis.