April, 1930 AIRCRAFT ENGINEERIN G 79 A Discussion of the Physical Aspects of Measuring Cabin Noises Quantitatively as a Preliminary to Taking Steps Towards Their Elimination By H. E. Wimperis, C.B.E., M.A., F.R.Ae.S. HE achievement of a reasonable degree of level, corresponding to an average speech seems just about as bad as that entering the silence in aircraft cabins is chiefly difficult intensity close to the mouth, of 1 micro-watt other from the source of sound under investiga- because of the amazing intensity of the per sq. cm., other intensities being measured tion. It is found unexpectedly easy to balance noise made by the airscrew and engine. There in decibels above or below this level. It is th e two, and different observers get much the is, in addition, the difficulty of excluding this bette r for the present purpose, however, to same results. The instrument reading, ir noise from the cabin without unduly adding avoid negative numbers of decibels and to decibels, is then taken, and that is all. The to the weight of the structure. consider the problem in relation to the scale device can be used if desired by "masking,' already mentioned of 0 to 130 decibels. In i.e., finding out what instrument setting pro Th e combined problem, difficult as i t is and must be, has been made in the past much more this scale the various common sounds can be duces a noise which can only just be heard ir so by the absence of any quantitative measure grouped as follows :— one ear when the noise under investigation is either of the trouble to be dealt with or of the entering the other; according to Davis, these Decibels. Noise. effectiveness of remedial measures. Fortu readings are about 20 to 30 decibels less thar 10 Rustle of leaves. nately, however, physical science has in recent th e others, and the method is of service when 20 Whisper. unusually intense sounds have to be investi years afforded a means of making such measure 40—70 Usual range of speech. gated. ments which is capable of use even in the 70 Busy street. difficult conditions of the ordinary aircraft Of course, the best way to realise quietude 70 Corridor of railway train. cabin. Recent work in the field of telephony in an aeroplane cabin is to avoid producing 90 New York subway. has provided also a convenient system of units loud noises just outside it. So it is desirable 100 Aeroplane. in which to state results. The new unit is first to see how quiet the airscrew and engine I t will be seen that if the noise in an aero th e decibel. can be made, and then so to build the cabin plane is to be reduced to that in an ordinary I n his " Text-book of Psychology " William as to reduce the effect of the residual noises railway train (and no less reduction would be James states Fechner's law in the following t o the desired low limit. form : "T o get equal positive additions to the If one pictures a cabin wall built of a sensation, one must make equal relative addi thin homogeneous substance, and the cabin tions to the stimulus." Put into mathematical closed and empty, one is inclined to expect a language, this means that if the sensation It will be remembered that in the decided reduction of sound intensity inside it January issue of AIRCRAFT EN increases by arithmetical progression, the pro compared with that outside. This, however, GINEERING an account of some ducing stimulus must be increasing by a is an illusion, since it would be just as noisy early experiments on the sounds geometrical one (e.g., Prof. Perry's " Compound inside as out. If the walls be hard and reflect emitted by aeroplanes was published, Interest Law ") ; or, in even more familiar ing, any sound that does penetrate will by and the subject dealt with at some terms, the sensation is proportional to the multiple interior reflections rise to the same length editorially. It is common logarithm of the stimulus. level as tha t outside; if to avoid this internal knowledge that the matter is now for the first time being adequately increase a thin and soft substance is used for tackled by the Aeronautical Research th e walls, so much th e more sound will penetrate A New Unit of Measurement Committee, at the National Physical from outside, and precisely the same results Telephone engineers have lately made use of Laboratory, and under the Director of as before will be arrived at. this law and have produced the unit known as Scientific Research, at Farnborough. the bel, and its tenth part, the decibel. Two Mr. Wimperis here details some of All Walls Equally Bad sounds are said to differ by N bels when the the preliminary physical methods of This is a somewhat startling conclusion. All discovering the chief sources of the logarithm (to base 10) of the ratio of the sound walls lead to th e same noise. The pat h of escape trouble, which should prove the first energies in each is equal to N. lies in the consideration that even when hard steps in what we hope will prove a N = log E /E . reflecting walls give an embarrassing accumula successful campaign to abate this 1 2 nuisance. tion of internal reflections, these reflections will Actually the bel is an inconveniently large be enormously reduced by the presence of cloth unit, and the decibel, only one-tenth as great, ed passengers, luggage, carpets, hangings, and is that commonly used. The zero of the scale th e like ; furthermore, some possible wall sub is taken as the degree of sound which is just satisfactory) a saving of about 30 decibels must stances turn a part of the sound energy into audible to the ordinary healthy ear; this is be made. heat, and so diminish the noise, and lastly, it called the threshold of audibility, and it corre I n setting about this saving the first need is is possible without too great an increase' of sponds to an air pressure of 1 millidyne per a n instrument for measuring his achievement structure weight to build a composite wall of sq. cm. It so happens that a change of sound as the investigator proceeds. And seeing that which the outer surface is designed to reflect intensity of one decibel is jus t about the smallest a difference of 30 decibels means a ratio of as much as possible and the inner surface to change the ear can notice. If the intensity is sound energies of 103, or 1,000 to 1, it is clear absorb as much as possible—in this way 10 steadily increased the limit is reached a t about tha t a very highly sensitive device is not needed decibels can easily be saved. 130 decibels, when the noise is too painful for in the first instance. Incidentally, it is useful hearing to go further. These limits are If the object be to save 30 decibels, the t o remember that halving the amount of noise naturally linked with the frequency of the saving of 10 by the use of suitable cabin walls, only leads to a decrease of 3 decibels; that note ; it could not well be otherwise, seeing leaves 20 to be recovered by a reduction in doubling the distance only leads to a reduc tha t frequencies below 20 and above 25,000 th e combined engine and airscrew noise. tion of 6 decibels; and that if a noise of 20 are not usually heard by the ear a t all. Airscrew noise probably chiefly* depends on decibels is removed from a combination of airscrew tip speed being as much as 40 decibels In the first attack on the aircraft problem noises producing 30 decibels, there will still higher a t a tip speed of 1,000 f.s. than a t 600 f.s. the investigator can deal with the mixture of be 29½ decibels left even when the 20 have If half of this saving could be made and be frequencies which make up aeroplane noise been removed. So that the clearance, if it coupled with a similar saving in engine noise without attempting to disentangle the mixture, ma y be so termed, must aim at the high- th e first part of the problem would be solved. just as photometric work can be done with a intensity noises and leave the lesser ones for I t must be remembered that engine silencing white light made up of a grea t mass of different subsequent treatment as necessary. is not worth while unless the airscrew noise wave-lengths. Telephone engineers adopt also An instrument suited to the needs of this can be vastly reduced, and that it is little use the convention of defining a zero-intensity problem is found in th e Barkhausen Audiometer. t o make silent airscrews whilst the engine is This instrument produces an adjustable noise so noisy. If both together, equally noisy, * An extract from a lecture on "Some Physical Problems in Aeronautical Research" delivered by the Director of Scientific in a telephone car cap, which is placed on reach a noise level of 100 decibels, the level Research, Air Ministry, to the Mathematical and Physical one ear, and the instrument setting is varied will still be 97 decibels even when one of them Society of the Imperial College of Science and Technology on March 11, 1930. until the noise it is producing in the one car is entirely suppressed.
Aircraft Engineering and Aerospace Technology – Emerald Publishing
Published: Apr 1, 1930