Study of the comfort aircraft norms

Sarah Jibodh-Jiaouan 1

Laboratoire Vibrations Acoustique (LVA), INSA-Lyon, 25 Av. Jean Capelle – 69621 Airbus, 26 Chemin de l'Espeissière, 31300 Toulouse, France Etienne Parizet 2

Laboratoire Vibrations Acoustique (LVA), INSA-Lyon, 25 Av. Jean Capelle - 69621 Villeurbanne, France Guillaume Osmond 3

Airbus Operation SAS, 26 Chemin de l'Espeissière, 31300 Toulouse, France

ABSTRACT In long range aircrafts, one of the main reasons of discomfort to the passengers is the vibrations and the noise induced by the rotation of the engine. This induces a significant level of single-frequency excitations (both for noise and vibration). The frequency (located between 25 and 55 Hz) and the amplitude of these excitations vary depending on the flight parameters. This study investigates the influence of the frequency and the level of vibration and noise stimuli to discomfort. The aim is to build a global discomfort model for acoustics and vibration. Subjective experiments were carried out in the laboratory, on a test bench allowing to control the vibration and noise stimuli. Volunteers separately assessed the vibration, noise and global discomfort. Foot, seat, backrest and armrest accelerations were measured for each participant and each stimulus. The relation between noise or vibration discomfort and excitation level and frequency was determined.

1. INTRODUCTION

Aircraft customers are very demanding on cabin noise and vibration comfort improvement. The aviation industry need to have a full understanding of the noise and vibration discomfort of the passengers in order to make their aircrafts more comfortable and to demonstrate the comfort of their planes to their customers. Acoustic and vibration comfort become an important marketing argument. The engine rotation is a significant source of noise and vibration in the cabin aircraft. At the moment noise discomfort and vibration discomfort are evaluated independently. Vibration ride comfort is estimated with the ISO 2631-1 standard [1] developed for any transportation and should be slightly adapted for each application. Studies at Airbus shows that this norm is sometimes not consistent with the crew or passenger complaints. 1 sarah.jibodh-jiaouan@airbus.com 2 etienne.parizet@insa-lyon.fr 3 guillaume.osmond@airbus.com

The aim of the present experiment is to study the influence of frequencies and amplitude on the passengers’ discomfort in order to improve the passengers’ experience in the cabin, especially close to the engines. In addition, the goal is to build a global discomfort model. 2. METHODS

2.1. Participants

Thirty-two people participated in this experiment including ten female and twenty-two males (mean age: 36.5 ± 9.8 year; mean height: 172 ± 8.0 cm; mean weight: 70.1 ± 12.0 kg). They were recruited inside Airbus and they come from different trades and areas of expertise. Participants and their labor doctors had to sign a certificate attesting their good health (no recent fractures or surgery, no retinal problem…) beforehand.

2.2. Stimuli

Stimuli construction was extrapolated from analysis using existing vibration and acoustic flight measurements performed in several cabin aircrafts of the Airbus fleet. The stimuli represented only the cabin vibrations and noise generated by engine rotation.

The nine vibration stimuli were composed of a mono-frequency vertical excitation. Three frequencies (25, 35 and 55 Hz) and three amplitudes (0.015, 0.025, and 0.040 g) were selected in order to address several aircraft types. The vibration stimuli duration was thirty seconds with around 4 s of fade-in. The nine acoustics stimuli were composed of cabin background noise (88 dB SPL - 73dB (A)) and a pure tone. Three frequencies (25, 35, and 55 Hz) and three amplitudes (75, 85, and 95 dB SPL) were selected for the pure tone covering the current and future fleet. The acoustic stimuli duration was twenty-five seconds with 10 ms of fade-in and fade-out. In total, there were twenty-seven combinations of vibration and acoustic stimuli because a vibration stimulus could only be generated with an acoustic stimulus of same frequency.

2.3. Apparatus

The vibration was generated by a six degrees-of-freedom vibration shaker (Airbus Cube in Toulouse), but only the vertical translational vibrations were used in this study. The top of the platform is approximately 150 cm x 150 cm, two rows of two economic class seats were fixed. The volunteers were placed on the right seat of the back row one by one. For each participant, each vibration stimuli was recorded by five triaxial accelerometers: two pads attached to the seat and the backrest, one accelerometer fixed at the ground of the platform and one on each armrest. All vibration parameters were sampled synchronously at 800Hz using a Siemens Scadas Mobile front-end.

The noise was generated using Matlab with an RME Fireface UFX II soundcard, two subwoofers (Genelec 7380 AP) and two loudspeakers (DAS Audio - Artec 508). The acoustic setup was calibrated once before the experiment. A microphone was placed between the theoretical positions of the passengers’ ears just as the in-flight procedural measurements used in the company.

2.4. Procedure

The experiment was divided in three blocks: Global Discomfort (GD), Acoustics Discomfort (AD), and Vibration Discomfort & Body Map (VD&BM). The participants were submitted to the twenty-seven combinations of vibration and acoustics stimuli for the GD block, to the nine acoustic stimuli for the AD block and to the nine vibration stimuli for the VD&BM block.

The participants analyzed the global discomfort for the GD block, the acoustic discomfort for the AD block and the vibration discomfort for the VD&BM block. Based on ISO 2631-4, a continuous discomfort scale between 0 and 100 with a semantic reference was presented to the participants for the three blocks. They had to give a numerical value corresponding to their discomfort. In addition, for the VD&BM block, using a Body Map (Figure) the volunteers had to answers to two questions:

- “In which area(s) of your body do you feel the vibrations?” - “In which area(s) of your body do you feel particularly uncomfortable vibrations?” There is no indication on any appropriate number of areas.

Figure 1: Discomfort scale and Body Map

All the answers were given aloud after each stimulus and the experimenter saved them on a computer. In order to familiarize participants to the stimuli and the discomfort rating method, sample signals were played to participants at the beginning of each experiment. The apparition order of each block was compensated and the stimuli were presented randomly for each participant. It was indicated to the volunteers to have their back stuck to the backrest, their arms on the armrests, and to move as little as possible. Each session lasted approximately 1 hour including briefing and installation of the participant on the seat.

2.5. Analyses

This experiment is based on a 2 x 3 complete factorial design in AD and VD blocks and in a 2 x 3 x 3 factorial design in the GD block. Each block was analyzed with a Repeated Measures Variance Analysis (RM-ANOVA) in order to investigate the factors affecting the discomfort. These were followed by post-hoc T-Tests with Bonferroni corrections. 3. RESULTS

Four participants had to be removed from the dataset due to abnormal functioning of the measurement setup.

The average of the participants’ evaluation of the acoustic discomfort varied from 53 to 68 on the discomfort scale. This corresponds to discomfort level ranging from “A little uncomfortable” to “Very uncomfortable”. The average of the participants’ evaluation of the vibration discomfort varied from 28 to 58 on the discomfort scale. This corresponds to a discomfort level ranging from “Very little uncomfortable” and “Uncomfortable”. The average of the participants’ evaluation of the global discomfort varies from 48 to 74 on the discomfort scale. This is translated as a discomfort level between “A little uncomfortable” and “Very uncomfortable”.

Ratings of noise discomfort increased with pure tone level and depends of the frequency. The ANOVA showed a significant effect of the pure tone level ( F (2,54)= 25.47, p <0.001), the frequency

Extremely uncomfortable Very uncomfortable Unconformable Little uncomfortable Very litle uncomfortable Not uncomfortable sesssssssss

( F (2,54)=9.07, p <0.001), and a significant interaction between the level and the frequency ( F (4,108)=10.87, p <0.001). Post-hoc T-Test confirmed a change in the rated noise discomfort between 25 and 35 Hz ( p =0.012) and between 25 and 55 Hz ( p <0.001) but there was no change between 35 and 55 Hz ( p =0.765). Post-hoc T-Test confirmed a change in ratings noise discomfort for the pure tone level between 75 and 95 dB SPL ( p <0.001) and between 85 and 95 dB SPL ( p <0.001) but there was no change between 75 and 85 dB SPL ( p =1.000).

Ratings of vibration discomfort increased with vibration levels and was dependent on the frequency. The ANOVA showed a significant effect of the vibration level ( F (1,27)=89.84, p <0.001), the frequency ( F (2,54)=20.6, p <0.001), and a significant interaction ( F (2,138)=4.74, p =0.010). Post- hoc T-Test confirmed a change in the rating of vibration discomfort between the three frequencies (25 – 35: p <0.001; 25 – 55: p =0.013; 35 – 55: p =0.002) and between the three vibration levels (0.015 – 0.025 | 0.015 – 0.040 | 0.025 – 0.040: p <0.001).

> 70 ~ 65 ~ = g — 8 a 60 ~ 55 ~ 50 ~ r 1 75 85 95 Pure tone level (dB SPL)

Ratings of global discomfort increased with pure tone and vibration levels and was dependent on the frequency. The ANOVA showed a significant effect of the pure tone level ( F (2,54)=9.62, p =0.001), the vibration level ( F (2,54)=48.72, p =0.001) and the frequency ( F (2,54)=31.46, p =0.001) and an interaction between the frequency and the pure tone level ( F (4,540)=2.70, p =0.030).

2 2228 88 3 8 o}woosig uoNeIaIA, 2 & 0.025 0.040 0.015 Level (g)

Frequency

Figure 2: Effects of amplitude variations on each frequency on acoustic discomfort and vibration

discomfort.

Figure 3: Effects of acoustic amplitude variations on global on each vibration level on global

discomfort for each frequency.

4. DISCUSSION

The results show that in this study there were more variations in the evaluation of vibration discomfort and global discomfort than acoustic discomfort. These results also show that acoustic discomfort and vibration discomfort depends on the frequency and the magnitude. There is a significant impact on vibration discomfort at all frequencies and all vibration level. There is a higher acoustic discomfort at 35 Hz and 55 Hz at 95 dB SPL. In addition, the results show that the global discomfort is a function of both acoustic and vibration discomfort.

5. CONCLUSION

The experiment investigated independently and simultaneously vibration and acoustic discomfort cause by engine rotation. This study provides a better understanding of the interactions of acoustic and vibration on the passengers’ discomfort. Further analysis will be done in order to find a global discomfort model. 6. REFERENCES

1. ISO 2631-1, Mechanical vibration and shock – Evaluation of human exposure to whole-body

vibration, Standard, ISO , 1997. 2. Griffin M.J., Handbook of human vibration, Human Factors Research Unit, Institute of Sound

and Vibration Research, The University, Southampton, U.K., 1990.