Thermal Comfort and Climatic Conditions: Temperature, Humidity, and Air Speed, Monografías, Ensayos de Ciencia de materiales

¡Descarga Thermal Comfort and Climatic Conditions: Temperature, Humidity, and Air Speed y más Monografías, Ensayos en PDF de Ciencia de materiales solo en Docsity! 190 9 Ergonomics and Packaging 25 in o Fig. 9.37. Position in the lateral view of the eyellipse Tf the back angle is not 25 deg, the position obtained above for L40=25 deg, must be modified in the side view. The X and Z displacement can be found using the following expressions: X = -9.331288 + 0.404789L40 — 0.0012611L40?, Z = 1.067621 + 0.0156987L40 — 0.00233471-40?, where L40 is expressed in deg (deg) and the displacements X and Z in inches (in). If the back angle is smaller than 25 deg, the eyellipses are moved forward and upwards. Conversely, if L40< 25 deg, the eyellipses are moved rearward and downwards, In the plan view (the plane xy of the vehicle reference) the xx line of reference frame of the eyelli t at a distance from the vehicle centerline given by 0.85W7 +0.075W3 where W7 is the distance of the steering wheel axis from the centerline and W3 is the width inside the vehicle 10 in (254 mm) above the R point. In any case the xx reference line should be located no further inboard than W20+1.1in.W20 is the distance between the R point and the vehicle centerline. This positioning is valid for vehicles with individual driver's seat, as it is for most current cars. As a first approximation, the xx reference line passes through the R point in plan view. SAE 3941 indicates the procedure to locate the eyellipses also in the case of bench seats, buses and trucks. s sea Critical aspects of the approach based on the eyellipses Eyellipses have been defined by considering seats provided with longitudinal and back angle longitudinal adjustment of the size and location of the eyellipses. The drive seat of modern ca subcompact and compact), is djustments. Their construction shows the important role of the s, even of the smallest cl: 191 aflect virtual manikins and provided with vertical and cushion slope adjustment. Th the definition of the eyellipses: One approach is to use repeat the procedure leading to the definition of the eyellipses and adapt them to the new conditions. Measurement of the ambinocular field of view Fig. 9.38 shows the procedure indicated by SAE J1050a Recommended Practice to determine the ambinocular field of view which, in the lateral view, is delimited between a line at 45 deg upwards and a line at 60 deg downwards. Viewpoints C and D are on the same plane as points F and G that are aligned in the vertical direction. In the plan view the ambinocular field differs depending on the possibility of head turning. Fig. 9.38 shows the construction without head turning. The right limit of the field of view is a sight line that starts from the left eye, the left limit is a line that starts from the right eye. These two lines have a maximum angle of 30 deg from the straight ahead view which corresponds to the maximum angle allowed by the rotation of the eyes without head turning. The construction is as follows e draw a line tangent to the right eyellipse at a maximum of 30 deg to the left, the tangent point is A. + Draw a line tangent to the left eyellipse at a maximum of 30 deg to the right, the tangent point is B. The ambinocular field of view without head turning o, has a maximum am- plitude of 60 deg and is between the sight lines passing through A and B. This angle may be less in the presence of obstructions (for example the A-pillars). Fig. 9.39 illustrates the procedure to determine the ambinocular field of view (ap) with head turning: + Drawa line tangent to the right eyelli point is G. at 30 deg to the left, the tangent + Identify point H on the left eyellipse by drawing a line from G parallel to the yy axis. Points G and H should be at about 60 mm apart. + Identify the hinge point J along the perpendicular to the midpoint of seg- ment GH at a distance of 3.88 in (98.6 mm). Point J represents the inter- section with the horizontal plane of the head axis of rotation. Points G” and H represent the eye points after head turning towards the left. + For a head rotation in the opposite direction (right), the same procedure must be repeated starting from a point G located on the left eyellipse. The maximum allowed rotation of the head about point J is 60 deg in both the left and right directions. 194 9 Ergonomics and Packaging D= 665 mm Fig. 9.40. Direct field of view following the European Directives. The longitudinal and vertical position of the visual points V1 and V2 must be modified as function of the back angle (L40). between a horizontal plane passing through V1 and a pyramidal surface with the top at point V2 (Fig. 9.41). The pyramid has three faces at 4 deg relative to the horizontal. Part of the steering wheel can exceed the lower limit provided it stays below a plane with a 1 deg downwards slope from point V2. The sloping and short hood used in modern cars usually does not interfere iew. In certain cases the belt line can introduce an obstruction pecially that on the opposite side to the driver (right door with the driver to the lateral vi for left hand driv A-pillar obstruction Fig. 9.42 shows the procedure indicated by the SAE J1050a [17] recommended practice to identify the binocular obstruction angle of the A-pillar. The reference closer one to the left eyellipse. This section should be found odure considering the lateral and the longitudinal views. ot field of view with eye movement, lering the points of view on the ion falls outside the direct field of view, the a procedure to rotate the head and eye so as to ion within the binocular field of view: ion is the cross sec by an iterative pro Tf the reference e only the binocular o eyellipses. If the rel recommended practice include the reference c tion falls in the dire truction is identified cons TO section AA of the pillar with a horizontal s through point B as 1. Side view: find the reference cros: plane (parallel to load standard B). This plane pa: 9.10 Visibility 195 Table 9.15. Corrections to the position of points VI and V2 of Fig, 9.40 as function of the back angle. Positive x to the rear, positive 2 upwards. L40 Az Az L23 Az Az (deg] | [mua] | [mm] || (deg] | [mon] | [mmm] 5 -186 | 28 23 -18 5 6 -177 | 27 4 -9 3 7 -167 | 27 0 0 8 -157 | 27 9 3 9 -147 | 26 17 5 10 -137 | 25 26 -8 11 -128 | 24 34 -11 12 -118 | 23 13 -109 | 22 14 -99 21 15 -90 20 16 -81 18 17 -72 17 18 -62 15 -14 -18 19 13 37 20 1 38 21 -35 9 39 115 -48 22 -26 7 40 123 -52 tion is found defined in the following. As a first approximation, this cross by cutting the pillar with a horizontal plane passing from the intersection of the eyellipse and its major axis. 2. Plan view: from point B, draw a sight line tangent to the rightmost point E of the e ction. The point B is the point of the eyellipse that is to the closest to left edge A of the « ¡on. Iterate the newly found point B on the side view and, if necessary, find a second approximation of the ection. reference cros 3. Find the point C on the right eyellipse such that the segment BC is parallel to yy- 4. Find point D at 3.88 in (98.6 mm) on the perpendicular to segment BO through its midpoint. Point D represents the hinge point for head move- ments. If the cross-section is all to the right of a line at 30 deg from point B, no head movement is required. In that case the binocular obstruction is the angle between sight lines AB and EC. 5. If part of the cross section is to the left of the 30 deg sight line from point B, the head must be rotated about point D until the sight line from B' to A forms a 30 deg angle with segment B'C”. 6. The binocular obstruction angle a: is between the sight lines AB” and EC”. 196 9 Ergonomics and Packaging Fig. 9.41. European directive about direct field of view. Only the A-pillars can exceed the field between the horizontal plane through V1 and the pyramidal surface with top at point V2. The three planes of the pyramid are at a 4 deg downwards slope. The binocular obstruction angle should not be larger than 6 deg. If the sight lines (AB' and EC' or AB and EC) tangent to the pillar converge (on the opposi part of the pillar relative the eyes) there is no binocular obstruction. In fac a certain distance from the pillar at least one eye is able to view again. Even if the driver has not a binocular view, he can appreciate the presence of obstacles, this justiñies the null obstruction. Although it would be possible to start from the eye points located to the rear of the eyellipses, this procedure considers points at the front part since this results in a higher binocular obstruction angle. Fig. 9.44 shows the requirements of the European Directive D77/649 and Ds1/643, D88/366, D90/630 to determine the binocular obstruction angle of the A-pillar: 1. From point R locate points Py, Pa, P.n given by Tab. 9.16 which represent the hinge points of the head when the driver looks out of the vehicle to the left (Py) or to the right (P») on a horizontal plane. 2. If the horizontal P;n must be correc t travel is larger than 108 mm, the position of Py, Pa, od as indicated in Tab. 9.17. 3. If the back angle is not 25 deg, the position of P 1, Pa, Pm must be corrected as indicated in Tab. 9.15. tions of the 4. Side view: starting from point P,n, find two reference cross pillar as indicated in Fig. 9.43. The c ions Si and Sa include the structural part and the portion of the glass with optical characteristics not complying the requirements about tansparency all non transparent part > eye points El, E2, E3, E4, located on the les of 65 mm (Fig. 9.43). 5. From points Py and Pa find the edges of two triangles with all s Table 9.17. Corrections to the x coordinates of points P, P2 for longitudinal seat travel larger than 108 mm. horizontal seat travel L23 | Ax [mm] 108 + 120 mm -13 132 mm -22 145 mm -32 4 158 mm -42 > 158 mm 48 9.10.4 Indirect Visibility The field of view of the driver is increased in the backwards direction by means of mirrors. In Europe the reference directives are D71/127, D77/649, D81/643, D88/366, D90/630. Fig. 8.38 shows the monocular, binocular and ambinocular fields of view through a mirror. Points OD and OS represent the driver's eye points which are at a distance of 65 mm from each other, at 635 mm (25 in) on the vertical of point R. In th >, this position is not affected by the back inclination and horizontal seat travel refer to an ambinocular view probably because when looking, s rather difícult to appreciate the distance that from an obstacle, Tt is therefore accepted that part of the vision is of monocular type so that the able to recognize simply the presence of the obstacle. For class My vehicles the European directive requires the following; The directive in a mirror it driver is + Inside mirror, Fig. 8.39 (top, to the left), the driver should be able to view in a rectangle on the ground 20 m wide and extending to the infinite starting from 65 m from eye points OD and OS. should view the + Outside mirrors, Fig. 8.40 (top, at center), the drive following rectangles on the ground extending to the infinite: — to the left, 2.5 m wide, starting from 10 m from the eye points. — to the right, 4 m wide, starting from 20 m from the eye points. Size and location of the mirrors The inside mirror should be of category 1. The size of the reflective part must be enough to inscribe a 4 cm height and a cm wide rectangle: 1 a=15 em: — py (9.4) 7 200 9 Ergonomics and Packaging r (expre found from the minimum and maximum radii Twin; Tmax at the centroid of the reflective part. ed in mm) is the mean radius of curvature of the mirror which can be ”= Fam ma > 1200 mm (9.5) The inside mirror should be completely above point V1 and at a minimum dis- tance of 350 mm from it in longitudinal direction. In order to reduce the « tortion of the reflected image, it is suggested to use a planar mirror whene possible, The outside mirrors should be of category TIL. The must be enough to inscribe a 4 cm height and a cm wide rectangle size of the reflective part 1 a=13 em — py (9.6) > where 7 (in mm) is the mean radius of curvature of the mirror. ”= Fam ma > 1200 mm (9.7) Additionally, it must be possible to inscribe a 7 cm vertical segment. In the plan view, the segment that connects the centroid of the refle to the center of segment OS - OD should be at less than 55 deg relative to the longitudinal direction x. Since the lower part of the outside mirror enclosure is less than 2 m from the ground, it should not increase the width of the vehicle of more than 20 cm. The width is measured without taking into account local tive part features such as handles or other small elements on the body surface. 10 Climatic Comfort Some knowledge regarding the physiology of the climatic comfort will be intro- duced in the following sections; at this stage it is useful to understand the mission of the climate control system [19][20]. irstly thermal comfort conditions will be identified as well as system param- s capable of describing them; secondly the human body metabolic activity as function of ambient conditions will be introduced and its consequence on heat exchange will be des At this point it will be possible to draw a thermal balance to design the climate control system prelimina Afterwards the design and test criteria for the most important parts of this stem will be introduced, referring to the explanation of their operation reported in Volume L. Volume 1 also reports information which is relevant as concerns the design of conditioned air distribution in the passenger compartment. ibed. 10.1 Physiology Outline Thermal comfort is defined by ISO 7730 Standard as expresses « ion for the thermal environment where a subject is exposed. The complexity of describing thermal comfort with suitable words compri also its definition. L. Morello et al.: The Automotive Body, Vol. 2: S springerlink.com O Springer stem Design, MES, pp. 201-238. ence + Business Media B.V. 2011 204 10 Climatic Comfort The thermal heat loss to the environment may be accomplished through the following modes of exchange': + skin sensible heat; e latent heat due to perspiration (Esw); + heat loss due to perspiration (Eq); e heat loss due to respiration (Ces); e latent heat due to the water contained in breath (Eres). The heat exchange of the skin includes conduction, convection and radiation; however it may be desc ¡ve contribution C and a radiation contribution R, taking the external clothing surface as reference. Physiology studies of the human body have demonstrated that no cooling- down, warming-up or evaporation occur if the body od to an ambient temperature within: ibed as the sum of a conve expos + 290-319 if the body is not covered by clothing; + 230279 if the body is clothed, in sedentary activity. The body in this temperature interval falls in a neutral zone and no physio- logical control action of the body temperature takes place. The skin temperature (t.,) and the internal temperature (+, condition at the following values: kin = 33,77C; e term =36,80. are in neutral For engineering applications the human body is modelled as a simple ad interactions with the environment are described with elementary pro in Fig. 10.1. ylinder as 10.1.2 Thermal Comfort Condition A subject considers the environment to be comfortable when thermally neutral, implying no sensation of being either hot or cold. In this situation, two conditions must be met. The first requires equilibrium between the internal temperature and the skin temperature. The second requires that the heat produced by metabolic activity must be equal to that lost by the body during its current activity. 1 The sensible heat derives from the temperature variation of a given mass due to convection, conduction and radiation or a variation of this mass, while the latent heat derives from mass state variation, as evaporation or melting. 10.1 Physiology Outline 205 Ambient boundary surface Ambient air CONVECTION RADIATION Skin EVAPORATION PERSPIRATION EN TA Clothing RESPIRATION N External surface Fig. 10.1. For engineering applications the human body is modelled as a simple cylin- der ad interactions with the environment are described with elementary processes. 206 10 Climatic Comfort Thermal comfort equation The description of the balance between energies in play was proposed by Fanger(22] as way of expressing the comfort condition using physical parame- ters, included in the so-called thermal comfort equation. With reference to the model in Fig. 10.1, the thermal balance equation could be the following; M-W = Qs + Qres = (C + R4 Esh) + (Cres + Eres); where: + M is thermal flux generated by the human body; e W is thermal flux exchanged with the environment; e Qsk is the thermal flux exchanged through the skin; e Qres is the thermal flux exchanged through respiration; e Cros is the thermal flux due to convection, in respiration; e Eres is the latent heat in respiration; + C is the convective heat exchanged through the skin; + Ris the radiation heat exchanged through the skin; e Es is the evaporation heat exchanged through the skin = Ersw + Ed. The terms of this equation often refer to the external surface of the bare body Ap, that can be inferred from the empirical Du Bois' formula: Ap =0, 292028978, where: e mis the mass of the subject; + his the height of the subject. The formulae of the following sections refer to the evaluation of the terms of the equation above. 10.1 Physiology Outline 209 In addition, two parameters describe the subject: e the metabolic index; + the clothing index. The following sections will be addressed to explain the parameters above. Air temperature Air temperature t, is measured around the subject but outside of its boundary layer. This temperature has influence on the heat exchanged by the human body because of air convection. Air temperature is usually measured by means of a dry bulb thermometer and, for this reason, is sometimes called dry bulb temperature. Mean radiant temperature Mean radiant temperature t, is the uniform temperature of a black cavity where oceupants would exchange the same amount of heat as in the real non-uniform ambient. It depends obviously on the temperature of all surfaces of the room enclosing, the subject and on any other surface the human body can receive heat energy by radiation thermal exchange. The difference between mean radiant and air temperature is positive, by defi- nition, when surfaces enclosing the subject are warmer than his skin, is negative if the opposite condition appli The precise calculation of mean radiant temperature involves many assump- tions on surface radiation and on multiple reflection within the enclosed space. Assuming that all surfaces exchanging radiation energy with the human body are black, a simplified equation allowing the calculation of the mean radiant temperature can be defined: A = Til) + Ti Fp2 +. where T; are the temperatures (measured in “K) of all surfaces radiating energy to the subject, while 77,_; are the view factors, that are coefficients defining the fraction of radiating energy of a given surface pertaining the subject in questi and satisfying the following condition: n 2 F i¡=1 i The evaluation of the view factors is the most delicate aspect of this calcula- tion. The direct measurement of the radiant temperature can be measured using a globe-thermometer. 1t is simply a dry bulb thermometer enclosed in sphere of 210 10 Climatic Comfort treated copper (black metalline) with diameter 150 mm and posse the same absorption characteristic of the human skin. As a demonstration of the importance of the radiating component in the thermal exchange of the human body, the actual adsorbtion coefficient of the human skin has been shown to be higher than any other substance including black metalline, As a consequence human being are very radiant temperature. sing nominally nsitive to any variation of the mean Air speed The air speed relative to the human body va plays an important role in evaluating the thermal comfort of a subject in an enclosed space with artificial climate because it influences the convective thermal exchange with the human body and is one of the control parameters of all air conditioning systems. In hot or humid weather, any motion of the surrounding air can increase the heat loss from the human body at the same air temperature. This fact is justified by two different mechanisms Tf the air temperature is less than the skin temperature, an increase in air speed increases convective heat exchange because a higher amount of fresh air vashes the skin. In the case that the ambient is moderately humid (between 30+85% of rela- tive humidity), the increased air speed will enhance also sweat evaporation by removing saturated air and replacing it with dryer air. The air speed in question is the average speed on the boundary layer between the ambient and the human body. Air humidity Air humidity ua refers to the amount of humid steam contained in a volume of air. At a given dry bulb temperature, the quantity of humid steam that can be adsorbed by the air before saturation and consequent precipitation is called absolute humidity. The saturation point also called dew point refers instead to the maximum amount of humidity at a given air temperature. The relative humidity (RH) is the ratio between absolute humidity and dew point humidity. Relative humidity is relevant to sweat evaporation. TF RH is over 80%, most of the sweat cannot evaporate and the air surrounding, the human body becomes quickly saturated. On the contrary, i£ RH is below 20%, mucous membranes become dry quickly with an increased risk of irritation and infection. Air humidity has a modes environments; nevertheless there are limits that should not be exceeded. This parameter is measured by hygrometers. influence on thermal comfort in air conditioned 10.1 Physiology Outline 211 Discomfort zone Comfort zone Fig. 10.2. Diagram identifying comfort and discomfort zones; the diagram is quoted with ta, the dry bulb air temperature, RH, the relative humidity and va the air speed. 214 10 Climatic Comfort room must exhibit the same skin temperature of wet skin 1. tsx and have the same percentage In summary, to measure the effective temperature, the reference room must r speed, the air temperature must be equal to the mean radiant temperature and its relative humidity must be 50%. have the same Metabolism The human body is a kind of chemical laboratory in continuous activity. Food, cal reactions that drink and other substances undergo a huge number of chemi together constitute the human body metabolism. Metabolic processes are basically oxidation which produce thermal energy; usion the potential chemical energy of food and drink (not including ubstances) and of other substances used ide the human body. thermal energy in is energy is the difference between the consumption of potential chemical energy and of useful work produced, and its effect will be to cause an increase in the body temperature, This energy is also called metabolic energy or energetic metabolism. The measurement unit for metabolism is MET, corresponding about to 58.15 W/m; it is the metabolism of a non working sitting subject divided by its body ce an average subject has a body surface of about 1.7 m 1 MET corresponds to about 100 W. The minimum value for metabolism is about 0.8 MET when the subject is sleeping and can increase up to 10 MET for heavy manual work or sport ac ties. Some average values for metabolism are reported in Tab. 10.1. Metabolism is inftuenced by the following factors: , a metabolism of . ag law metabolism decreases with increasing age with an almost proportional e sex: female metabolism is roughly 5% lower than male metabolism at the same age. Although the exact value of metabolism cannot be measured directly, the following values can be assumed for reference: 44 W/m? for a male and 41 W/m? for a female subject. Clothing Clothing Index CLO provides a measurement of the thermal insulation due to clothing: clothing reduces the body heat loss to the environment affecting the thermal balance. This index is a thermal resistance that is usually measured as a ratio with a reference value of 0.155 m22C/W, representing a typical situation of clothing suitable for a sedentary office activity. 10.1 Physiology Outline — 215 Table 10.1. Some values of metabolism measured in MET and W/m?. Kind of activity 1W/T | Met Bedded subject 46 8 Seating subject 58 LO Standing subject 70 LI Office work 70 12 Driving a car $0 14 Shop working 93 16 School teaching 95 16 House keeping 100 17 Walking at 2 km/h 110 19 Laying-down bricks 125 22 Gardening 170, Heavy housekecping 170, Dismantling with a pneumatic hammer | 175 Walking at 5 km/h 200 Wood cutting with a motor saw 205 Ice skating at 18 km/h 360 Dieging 380 Cross-country skiing 9 km/h 405 Wood cutting with an axe 500 Ruming at 15 km/h 550, In these conditions, the clothing index is CLO = 1; a bare body would have a clothing index CLO =0. The clothing index increases s when the clothing is suitable for use outdoors in the winter season, while decreases in summer season. Some values of CLO are given in Tab. 10.2. The overall value of the clothing index CLO; can be calculated by adding CLO) indices pertaining to each piece of clothing, as shown in Fig. 10.3, to the following formula: n CLO;= Y CLO,. This way of proceeding provides data that are usually sufficiently accurate; if more accurate data are necessary, the total heat re ssigned clothing s in a climatic cell. Also stuffing present in any piece of furniture contacting the human body, eg. ats ion and must be taken into account. ar can be measured using internally heated human dummi car , exerta S nificant influence on heat rej 216 10 Climatic Comfort Table 10.2. Some clothing indexes measured in CLO and m?"C/W. Type Description CLO | [m*C/M] Underwear underpants 0,03 0,005 boxer: 0,06 0,009 0,13 0,020 undershirts 0,09 0,014 Shirts ds 0,09 0,014 Tong s shirts 0,12 0,019 flannel long sleeved shirts | 0,30 0,047 as above, with roundneck | 0,34 0,530 Trousers shorts 0,06 0,009 long light 0,20 normal 0,25 fannel 0,28 Suits coveralls 0,50 S 0,30 Sweaters S 0,12 light 0,20 Tight with roundneck 0,26 heavy with roundneck | 0,37 Jackets light 0,25 normal 0,35 Overcoats regular overcoats 0,60 winter jacket 0,55 heavy winter jacket 0,70 Miscellaneous socks 0,02 stocking 0,10 light sole shoes 0,02 heavy sole shoes 0,04 boots 0,05 Seats wooden or metal 0,00 padded 0,10 padded with armrests | 0,20 0,032 10.1 Physiology Outline 219 2 cool wall 7 hotceiling * coldceling hotwall PROS : ] 5 0510152025300 0 510152025300 '0 510152025300 0 5 10152025300 At At Fig. 10.5. Some example of local discomfort caused by vertical and horizontal thermal gradients. For this reason it is more sensitive to the relative humidity than PMV and this sensitivity increases as the humidity in In summer the comfort zone relative humidity between 20% and 60%. In winter the comfort zone is characterized by ET” between 2022,920 and, again, relative humidity between 20% and 60%. orized by ET* between 22,8+26%C and Local discomfort Although a subject could be in a neutral comfort situation, sone parts of his body cold be expo lar conditions caused by unevenne parameters: This called local discomport. This kind of discomfort cannot be corrected by changing average air temper- ature or humidity, but acting locally. The most frequent causes for local discomfort are the following: d to parti ituation is + too hot or cold room walls; + too hot or cold floor; + asymmetry in radiant thermal flux; e air draft. Fig. 10.5 illustrates some examples of local discomfort caused by horizontal and vertical temperature gradients; the diagrams below e represent PPD as a function of the temperature differenc se situation image between a reference ion (dotted in the drawing) and one of the walls around the room in qu tion. The first three situations are relative to winter heating, whercas the last corresponds to summer cooling. 220 10 Climatic Comfort Vertical temperature gradient This effect is caused by the thermal gradient existing in closed room as a conse quence of the change of air density due to temperature, This phenomenon cau: a feeling of hot at the head and of cool at the feet. Experimental results report that a difference of 3C between head and foot is rated as discomfort by 5% of interviewed subjects. While a higher head tem- perature at the head that at the foot may generate dissatisfaction, the opposite situation is never rated as discomfort. Floor temperature A diferent floor temperature may affect foot comfort. It is inappropriate to consider floor temperature because the heat loss only affects l mfort; this heat loss depends on the thermal conductivity of the floor and shoes. It has been reported that sitting people al than standing people. ISO 7730 specification provides for a floor temperature in the range between 19290. pt a floor temperature 19C higher Radiation asymmetry This kind of discomfort arises when a part of the body is subject to a heat radiation and the radiation flux is not uniform. This situation can be described by radiant temperature, measuring the differ- ence of this temperature on the two opposite sides of a same plane portion. A hot ceiling with a cool window is a cause of higher discomfort than hot walls and a cool ceiling. Air drafts Discomfort occurs when an undesired local cooling is caused by a concentrated air stream. It should be noticed that the human body is relatively sensitive to air motion since it stimulates thermal sensors to convey signals to the brain. Many alert signals that are irrelevant independently may together cause high anmoyance. To describe air speed fuctuations, y T is frequently considered; it is defined as: where 9 is the standard deviation of air speed and va the average speed of the flow. Fig. 10.6 reports some empirical correlations between PP.D and air speed Va at a given air temperature t, with different values for the turbulence intensity 7. 10.2 Passenger Compartment Energy Balance — 221 v¿[m/s] PPD = 15% va[m/s] PPD = 25% 05 05 :=9/ =10/, 1=0 =10 /=2 DA: > DA: =40, D3 9 03: == 0.2. 0.2. 01 01 r q 18 20 2 24 2 2 30 18 2 2 2 2 2 30 taPC] taPC] Fig. 10.6. Some empirical correlations between PPD and air speed va at a given air temperature ta with different values for the turbulence intensity 7. 10.2 Passenger Compartment Energy Balance The function of the air conditioning system is to supply thermal energy at a controlled temperature in such a way as to improve thermal comfort conditions. The first step necessary is to understand which are the parameters that play a role in the thermal balance of the passenger compartment. The heat flows entering and exiting the passenger compartment can be s marized in a thermal balance equation where the thermal power introduce the climate system is compared with a series of not controlable terms. In the expression: Wimp = Wa + Wi +W, + Wan, e Wimp is the power introduced by the climate system; + W. is the power exchanged between the passenger compartment and out- side by means of convection and conduction; e W, is the power radiated through glass: + W, is the power introduced by passengers because of metabolism; + Wi is the power introduced by the powertrain because of the higher tem- perature of its parts. Each term of this equation is described in details subsequently. Furthermore some other parameters must be taken into account including temperature and humidity. 224 10 Climatic Comfort Table 10.3. Flow breakdown through the different boundary surfaces of the passenger compartment, hy winter heating. Part name | Percentage of total Roof 11 Classes 18 Floor 13 Dashboard 12 Doors 3 Remaining parts 38 10.2.2 Radiated Heat Particularly in summer, an additional amount of heat due to sun radiation is brought into the equation. ct of sun radiation depends on the surface and the color of the car ontribution is two-fold: The oncoming radiated energy through the and the energy absorbed and again radiated by the body. combined effect of these contributions, which depend on the relative po- n exceed 1,000 W/m?. sition of the sun and on weather conditions. 10.2.3 Passengers Metabolism The previous sections have shown how the metabolism is influenced by human aking into account the reduced amount of activity while riding or driving in a car, an average value of 100 W per passenger can be assumed for this part[23/[24]. 10.2.4 Powertrain Power This contribution depends on many factors, including the car power demand and the engine size; an average value could be set around 300 W. 10.2.5 Air Conditioning System As explained in Volume passenger comfort is provided by the air conditioning section of an HEVAC system to indicate the function of each component; another is shown in Volume 1. The heating subsystem includes a radiator R exchanging heat between the engine coolant and the air in the passenger compartment. The ventilation subsystem includes the vents to introduce a certain amount of air into the passenger compartment from the outside through AE and from 10.2 Passenger Compartment Energy Balance — 225 DP Fig. 10.8. Cross section of the HEVAC system, showing the ventilator V, the evapora- tor E and the heater R; the treated air can come from the outside or the inside through the vent MSR and cross the evaporator alone or also the heater, by means of the vent SM. the inside via AL. Air circulation is activated by ventilator V. The air can be used directly as it is or its temperature and humidity can be adjusted through the evaporator E and the heater R. The evaporator E is part of air cooling system including a condenser, a com- pressor and an evaporation valve, usually located in the engine compartment. Through these four components, a refrigerating fuid performs a thermodynamic eyole capable of reducing the temperature before entering the evaporator. A set of valves, that can be controlled manually or automatically, perform a number of functions: + a valve MSR allows the ambient or the passenger compartment air, de- pending on which exhibits conditions closer to the desired value, to be introduced into the HEVAC system ; + a valve SM allows the air to cross the evaporator (to redu and humidity) and the heater (to heat it, if necessary); s temperature + Valves SDI, SDS and DS allow the air to be distributed to the passenger compartment in the most appropriate way and directed towards the feet or head of the occupants or towards the windshield. Air speed can be adjusted by changing the voltage of the electric ventilator. 226 10 Climatic Comfort The thermal power introduced into the passenger compartment may be e. culated as follows: Wi; = Qpep(Tin — Tu), where: e (is the air mass flow treated by the HEVAC system, e pisits density, e cp is its specific heat, e Tin is the passenger compartment air temperature, e Ti: is the treated air temperature. Car HEVAC systems are usually designed as to have the treated air crossing the evaporator in any case in order to ensure that the humidity is reduced while setting the evaporator temperature to the appropriate value. Tf the air leaving the evaporator is too cold, its temperature will be increased by heating. The heater can work in two different ways: + by dividing the air crossing and passing-by the heater appropriately, as shown in Fig. 10.8; + by ensuring all the air crosses the heater, adjusting the coolant flow ap- propriately using a valve. 10.3 Hevac System Design and Testing This sections addre s the heating and cooling subsystems since air distribution has been already covered in Volume 1 10.3.1 Description The scheme of a complete HEVAC system, including the engine compartment components, is shown in Fig. 10.9. The cooling system pumps the heat from a cooler source, the passenger com- partment, to a hotter environment, the exterior, by exploiting an inverse ther- modynamic eycle, The heat transfer o cle including pr valvo VE) and a physi orator and conden The thermodynamic eycle can be represented on a plane where the horizon- tal axis represents the enthalpy h of the working fluid and the vertical axis the 10.3 HEVAC System Design and Testing 229 component is negligible. The working fluid state changes from humid vapor (1) to dry vapor (2). The absorbed heat must now be di hange feasi perature to a value which exceeds the temperature of the air of the outside environment. The compressor provides this function, sucking the low pressure vapor coming from the evaporator (point 2) and compressing it to a higher pri sure value (point 3). With this pressure increase, the temperature also incre: The compression work is provided by the vehicle engine, moving the compressor through a mechanical transmission, usually a V belt The 2-3 curve can be represented by means of an isoentropical curve by ne- glecting the thermal exchange with the environment. It should be noted that is convenient that the starting point of the compression line corresponds to a dry vapor state because fluid droplets could damage the compressor. The working fluid, now in a gas state, enters a second heat exchanger that enables the dissipation of the heat to the external environment, according to the curve 3-4, Again it is assumed that the pressure remains unchanged during this tra, formation; in a first phase the vapor is cooled down, before being condensed to the liquid state. At the condenser outlet (4) the working fluid is high pressure liquid. This liquid enters now the expansion valve VE so that its pressure is reduced; this pressure reduction is accompanied by cooling down to reach the conditions represented in the point 1; the transformation line is isoenthalpical and the fuid returns to its initial condition. > exchanged heats Q1 and Q> and the compression work L, are represented rresponding enthalpy variations on the (p, h) plane. A common refrigerating fluid used in HEVAC systems is R134a, as mentioned in Volume L sipated to the outside environment. the fluid tem- To make this heat e: 10.3.2 Cooling System Here reference is made to the detailed description of the components of this system provided in Volume 1. The evaporator has the dual function of extracting the heat from the passenger compartment air and reducing its humidity. In fact the heat exchanged with the air has a sensible part: MacpAt, and a latent part, coming from water condensation: Mo», where: 230 10 Climatic Comfort Xx Saturation curve Sensible heat Constant Latent humidity heat S 1 curve Y 2 *, bh l t, £, [*C] Fig. 10.11. Psychrometric diagram of humid air; the horizontal axis represents the air temperature ta, the vertical axis the vapor title 2. e MM, is the air mass flow rate crossing the evaporator; e cy is the air specific heat; e Atis the air temperature change between the evaporator inlet and outlet; + Mo is the working fluid mass flow rate; +». is the specific condensing heat. The transformation of the air can be represented with a psychrometric dia- gram. The qualitative diagram in Fig. 10.11 shows the vapor title « on the vertical axis and the air temperature t, on the horizontal axis. In this case the vapor title is the ratio between water vapor mass and air mass. This diagram shows the so called saturation curve, the locus of points with the highest vapor content at a a family of curves with constant relative certain temperature; in the same wa humidity UR can be built up. The transformation of the air cros curve such as 1-5-2 and includes two parts; the first part with constant humidity to the saturation point, and a second part where both temperature and humidity decrease. sing the evaporator can be identified by a 10.3 HEVAC System Design and Testing 231 The treated air has both a temperature lower than initially and a lower vapor content (12 < 21). Evaporators used in today's plates and fins. The working fluid enters from the top and flows through plates with both internal (fuid side) and external (air side) fins. Fins are necessary to increase exchanging surfaces and consequently improve exchange eficienc; For sake of s the evaporator is contained within a rectangle of no more than 300 mm long, with a depth of about 60 mm, usually with two rows of plates. The total exchanged heat is typically in the range 2 to 7 kW. ars are cross flow heat exchangers, with laminar 10.3.3 Heating System The scheme of the heating s em is also shown in Fig. 10.8 comprising; + the engine radiator; + the thermostatic valve T; + the water pump P; + the heater and, obviously, + the engine, The system includes two parallel cooling cireuits for the internal combus- tion engine generating a relevant quantity of heat. During engine warm up the thermostatic valve shuts off the circuit including the radiator in such a way that all the available heat is sent to the heater. In this way the engine coolant heats the passenger compartment air. When the cooling fluid reaches an appropriate temperature in the range between 88+90%0, the thermostatic valve opens and the radiator circuit is activated as the generated the heater can receive. me technology and has the nominally same size as the heat exceeds that whic The h evaporator. The design power is typically in the range 5 to 13 kW. 10.3.4 Design Example The diagram in Fig. 10.12 is the actual Mollier diagram for the chlorofluorocar- bon R 134a; with respect to the qualitative diagram shown in Fig. 10.10, this shows other curve families including; + constant vapor title « curves, ranging from .1 to .9, inside the limit curve, where clearly x =0 on the left and x= 1 on the right; + the constant temperature curves ranging from —60%C to 240% 234 10 Climatic Comfort Point 2: + p,=.29 MPa; e ho = 400 kJ/kg; + s = 1.7kJ/kg%K. Point 3: + p3= 1.49 MPa; can be found at the intersection of the pre the temperature of 55 %C and the i curve with sg = 1,7 kJ/kg"K, identifying an ideal adiabatic compression through point 2. The thermodynamic parameters of point 3 that Mollier diagram are: jure line corresponding to pa, at .entropi an be achieved from the + ha = 433 kJ/kg; e t3=61.5C. Point 4 is on the saturation curve at 55 "C and therefore: e h¿= 281 kJ/kg. The enthalpy h4 of point 3 is equal to hy of point 1, because the expansion 4-1 in the expansion valve is assumed to be isoenthalpic. Point 1 is characterized by: e. ti=0C; e hi = 279.4 kJ/kg. The cycle defined in this way is represented in Fig. 10.12. The cooling heat is given by: Que = ha — hi = 399 — 279.4 = 119.6 kJ/kg. The working fluid mass rate nec can be calculated using; sary to obtain the desired power of 5 kW F=P, / Q 12. to obtain: F = 0.042 kg/s. The compression work is: W = (hz — ha)F, corresponding to: W = 1.43 kW. 10.3 HEVAC System Design and Testing 235 The mechanica -fficient of performance COP is defined as the ratio between spent energy and transferred heat: P ho —hx corps =25M W h3— ha and therefore: COP=3,5. For a new value for the condensation pressure pg = 20 bar with the same cooling heat, the cycle will be modified as follows: Point 2: + p=2. bar; + ho =400kJ/kg; e 5 = 1.73 kJ/kg"K. Point 3: + p=20 bar; * ta =67.5%. Point 3 is identified by the intersection between the line of constant pressure Pa and the isoentropic curve s3 = 1.73 kJ/kg"K, corresponding to an adiabatic compression through point 3, corresponding on the Mollier diagram to: + hz = 439.5 kJ/kg; e t3= 75.20, Point 3 lies also on the saturation line at 67,5C and therefore: + ha =300kJ/kg. The enthalpy ha of point 4 corresponds to the enthalpy /h1 of point 1, due to the nature of the expansion 4-1 through the expansion valve. In point 1: e ti =0%C; + hi = 300 kJ/kg. Hence the new value for COP decreases to: COP = 2.44. 236 10 Climatic Comfort 10.3.5 Testing ion of an HEVAC system is usually repre The miss ented by two basic tes tem + the first evaluates the maximum performance of the air cooling sy cooling down the passenger compartment starting from an initial condition corresponding to a long stop in the sun in a hot country; + the second evaluates the maximum performance of the air heating system, ic. heating up the passenger compartment after a long stop in winter time in a cold country. Traditionally hot and cold weather conditions are assumed to correspond to a southern state of the United States and to a country of Northern Europe respectively. Both tests are usually performed in a climatic wind tunnel, where the air temperature, speed and humidity are controlled; solar radiation is simulated using a set of high power lamp: Typically in a climatic wind tunnel up to 100 km/h which is lower than in conventional wind tunnels for acrodynamic research, although high air temperatures t, > 40%C and sun radiation of 1.000 W/m? can be recreated. A second major difference between climatic and acrodynamic wind tunnels is that for climate testing the vehicle engine must be in a condition to work properly, being a significant source of heat and mechanical power; for this reason a chassis dynamometer (sets of braked rollers) is used to simulate road load. Climatic wind tunnels are also applied to identify operational conditions in hot weather for each component working at high temperature conditions and to verify cold weather performance including engine startability and drivability. The significant amount of energy needed to maintain the temperature at e treme values suggests the opportunity to develop specialized cold and hot cli matic tunnels, an artificial wind simulates the car speed Summer mission The car under test is monitored with thermocouples in various key positions in the passenger compartment; the climatic wind tunnel typically uses a tempera- ture of 432C+1 with simulated sun radiation on the car roof of about 900 W/m. Passenger dummies equipped with thermocouples sit in the car. Conditions are maintained until the temperature at the dummy head is in the range between 63-65%C. The engine is started and the air conditioning system is adjusted to maximum cool; external air valve is set to recycle. Then the car is driven according to a driving eycle with speeds between 30 and 90 km/h in calm air. The wind tunnel must assure the temperature of 432C+1 and a relative hu- midity of 30%-+3% for the entire duration of the test. An acceptable test output is achieved if the following temperatures of the dummies heads are reached: 11 Noise, Vibration, Harshness When traveling, all vehicles are subject to several dynamic excitations. The in- duced vibration and noise has a number of effects on the vehicle and its oceu- pants ranging from the integrity of the structure to the perception of comfort and driving performance. High levels of vibration and noise tends to reduce comfort, the outcome being increased fatigue of the driver that in the long term has an impact on safety. The dynamic excitation induces repeated loads on the structure that reduce its life due to fatigue, hence influencing the safety and reliability. Dynami acting on the tires have a negative influence on the longitudinal and side slip forces, reducing the accelerations that can be achieved during traction, braking and cornering. The result is again reduced safety. These are just a few examples of how the noise and vibrations influence several aspects that ultimately may affect safety. The vibrations that occur in a vehicle involve a wide range of frequenci ranging from below 1 Hz for maneuvering loads up to about 10 kHz for acoustic excitations. The effect does not only depend just on the nature and the intensity of the excitation, but also on the dynamic behavior of the structure and the acoustic cavity inside the vehicle that can amplify the effect due to structural and acous . Vibro-acoustic phenomena are usually classified considering the conventional frequency ranges. With reference to Fig. 11.1 the following frequency bands can be defined: excl loads os ic resonanc: + Ride (0-5 Hz) which includes the low frequency accelerations due to vehicle maneuvers and rigid body oscillations of the car body on the suspensions. L. Morello et al.: The Automotive Body, Vol. 2: System Design, MES, pp. 239-363. springerlink.com O Springer Science + Business Media B.V. 2011 240 11 Noise, Vibration, Harshness Engine: combustion Aerodynamic noise nosie, mechanical noise, intake noise Exhaust noise Exhaust system á Differential and noise Brakes back axle noise Gearbox o Rolling noise Frequency [Hz] Fig. 11.1. Main sources of vibration and noise and indication of the main conventional frequency bands used to classify the vibratory phenomena. + Shake (5-25 Hz) which includes connected to the vehicle chas the unsprung mas: the first resonances of the main subsystems s such as the engine and its suspension, and s on the tires + Harshness (25-100 Hz) which includes the resonances of the car body as a flexible structure. This frequency range represents a partial overlapping of the frequencies that are perceived as vibrations with the noise. High intensity acoustic vibrations included in this range are sometimes perceived by the car as pre this is usually referred to as boom. jure variations e Noise (>100 Hz). For frequencies higher than 100 Hz the vibration is per- ceived by the ears as noise. The vibration feeling is significantly attenuated and can be barely appreciated by the tactile senses, Hence the band above 100 Hz is usually referred to as noise. The perception of noise is due to air pressure variations within the car. The noise may enter the interior of the vehicle through apertures; alternatively ex- terior noise may cause vibrations of the panels that form the vehicle body pro- ducing pressure variations inside the vehicle and hence interior noise, Depending on the mechanism at the base of the propagation, it is common to refer to two different vibration paths: 11.1 Sensitivity to Noise 241 + Structure born: : the vehicle subsystems transmit dynamic forces to the vehicle body directly by means of the connection points and structural interfaces. This induces vibrations in the structure and in the car body panels. The structural vibrations are transmitted to the air that surrounds the occupants inside the vehicle thus causing noise. For example, the engine vibrations are transmitted by the engine suspension to the chassis; the vibration of the chassis causes the car body panels such as the firewall and the floor to vibrate, therefore inducing acoustic pressure variations inside the vehicle and hence noise. + Air borne: The pressure waves that propagate outside the vehicle induce vibrations in the car body panels that, in turn, induce pressure waves in- side the vehicle, The vibration of the engine block and covers, for example, jure variations in the engine compartment; the vibration in- duced in the panels surrounding the engine is transmitted to the inside of the vehicle as noise, This example illustrates an air borne path including, also a structural element; instead, in some cases the existence of holes or direct air connection between an enclosure and another allow a direct trans- mission path through the air. The low attenuation commonly experienced in this latter case suggests the need to minimize this kind of transmission path already in the design stage. In all cases, the final phase of the noise transmission is in the air around and inside the cars of the oceupants. Distinguishing between air borne and structure borne noise is then a matter of considering the transmission between the source and the panels that surround the interior of the vehicle As a first approximation, structure borne transmission can be considered to dominate below 500 Hz while the air borne becomes increasingly important above 1000 Hz. 11.1 Sensitivity to Noise Macroscopically the human ear can be divided into three main parts (Fig. 11.3): e outer car— the visible car is a flap of tissue that is also called the pinna; + middle car — constituted by the auditory canal, the tympanic membrane (elastic membrane that separates the external auditory canal from the tympanic cavity) and from the small bones (malleus, incus, stapes); emi e inner ear — formed by three circular canals (devoted to the balance ) and from the cochlea, connected to the cochlear nerve. ns The outer ear collects and concentrate: the pressure waves to the auditory canal. The reflections of the sound on the convolutions of the flap produce a first 244 11 Noise, Vibration, Harshness Stapes LESS window Basilar membrane > A a Round window Fig. 11.4. Scheme of the cochlea. Fluid Stapes Oval window Cochlea (unwrap) lc (middle ear) E ES L£ * SS NJ] Round window Basilar membrane Fig. 11.5. The development of the cochlea. if it is superimposed with one of larger intensity, even if at a slightly different frequency since the fainter sound is “masked” by the louder one, even through the frequencies are different. fasking demonstrates a logarithmic dependence on the frequen At fre- ated as a different quencies below 500 Hz, a low intensity noise can be appre tone if its frequency is relatively closer than that needed to di frequencies that are both above 500 Hz. Fig. 11.7 shows the intensity of the noise that can mask a fainter one. The masked noise correspond to the minimum of each curve (shown with a dot). Each the intensity of the noise that is necessary to produce the s output from a given segment of the basilar membrane. The minimum of each curve correspond to the maximum sensitivity of that portion of the membrane. Due to the logarithmic scale, the frequency amplitude of each curve increases with increasing frequency. Because of masking, at least for sinusoidal signals, the perceived tone (or cern two different curve indicate me height) corresponds, to the minimum of the curve associated with a given seg- ment of the basilar membrane. Referring to Fig. 11.7, a 900 Hz signal will be perceived by the segment with maximum sensitivity at 1 kHz as a 1 kHz noise. 11.1 Sensitivity to Noise 245 Malleus hammer) ( ) Incus Stapes ¿Oval window Basilar membrane Tympanic Helicotrema membrane 5 10 15 20 25 30 de x [mm] sensitivity 20 25 30 y [mm] Fig. 11.6. The sensitivity of the basliar membrane as a function of the dista the oval window. For high frequencies (8 kHz o 6 kHz) the maximum s reached close to the oval window; for low frequencies (0,3 kHz) sensitivity is at the end of the cochlea (elicotrema). from ity is maximum The auditory response is not only function of the frequency and amplitude content, but also of the temporal sequence of the signal essentially due to the reaction times of the functions involved in the hearing system. The subj reaction tim os induce a “temporal masking” , such that a sound of finite duration masks the perception of sound events if they are separated by a too short a time interval. Similarly to frequency masking, temporal masking « shaped dependence on frequency. so exhibits a bell Intensity measurí The aim of the following section is to summarize the main physical quantities used to measure the sound intensity. Sound intensity level This is the power per unit surface that fows through a given area. At a given time + its value is , La(t) = p(t)5(t), (11.1) 246 11 Noise, Vibration, Harshness 1 y Y 0.1 1 10 f [Hz] Fig. 11.7. The level of signal necessary to mask a weaker signal. The masked sound corresponds to the minimum of each curce (marked with a point). ul +] + wk is the speed vector of the fluid. s proportional where p(t) is the pressure, (t) Due to the Euler equation, the to the pressure gradient (Vp) 1 d=--Vp; (11.2) p for a unidimensional propagation along the coordinate r, with speed u, Du 10p 2 __ 2. (11.3) Dt por To understand the physical meaning of the Euler equation, it is useful to consider a piston that moves in an cylinder opened at the end opposite to the piston. If the piston travels with constant speed, at steady state the pre tant in all the fluid column pushed by the piston. Conversely, if the piston accelerates, the fluid in contact with is compressed between the piston and the rest of the fuid column, that will tend because of its inertia to maintain a constant speed. The result is a pressure gradient, the resulting pressure wave propagating in the fuid column with the speed of the sound. Similarly, the accelerations perpendicular to the surface of a vibrating plate produce pressure waves that propagate around it. The speed u of Eq. 11.3 can be computed as the integral of the acceleration Du /0t: u= Za (11.4) por sure i Table 11.2. Correspondanci columns indicate the boudari 11.1 Sensitivity to Noise 249 cies in Hz and in bark. The first two h interval. Columns 3, 4, 5 and 8, 9, 10 indicate the central frequency of each band (f.) and the frequency interval (fa). z RT TV Año + TR VR TZ ZTo Bar | [4 | (Hg | Barx] | (1 || (Bar | (1 | (| [Bar | (Hi) 0 0 12 1720 50 0.5 100 1850 12.5 280 1 100 13 2000 150 15 100 2150 13.5 320 2 200 14 250 25 100 2500 14.5 380 3 300 15 2700 350 3.5 100 2900 15.5 450 4 400 16 3150 450 4.5 110 3400 16.5 550 5 510 17 3700 570 5.5 120 4000 17.5 700 6 630 18 4400 700 6.5 140 4800 18.5 900 7 770 19 5300 840 75 150 5800 19.5 1100 8 920 20 6400 1000 | 8.5 160 7000 20.5 1300 9 1080 21 7700 1170 | 9.5 190 8500 21.5 1800 10 1270 22 9500 1370 | 10.5 210 10500 | 22.5 2500 1 1480 23 12000 1600 | 11.5 240 13500 | 23.5 3500 12 1720 24 15500 1850 | 12.5 280 250 11 Noise, Vibration, Harshness Intensity Perception The sensitivity of the human ear is not constant over the nominal range 20 and 20.000 Hz; instead the sensitivity is relatively low at the extremes of the range (below 100 Hz and above 10 - 15 kHz) whereas it reaches a maximum in the intermediate frequency range. urves of Fig. 11.9 show, for each frequency, the sound intensity (or, more ly, the sound pressure level - SPL, in this case) corresponding to the same intensity perception. These curves are usually referred to as “isophonic and enable the perceived intensity to be quantificd with a measurement alled phon. At a frequency of 1 kHz, 40 phon corresponds to a sound re level of 40 dB. sophonic curves show that the maximum sensitivity of the human ear in the range 3 and 4 kHz, whereas the sound pri ry to give a certain feeling is lower than for higher or lower frequenc For high intensity (8090 phons), the curves are relatively flat compared to the low intensity curves (10-20 phons). Considering the curve at 20 phons, for example, at 3 kHz 15 dB are enough to produce the same feeling as 75 dB at 31 Hz (75-15=60 dB difference). Conversely, considering the 90 phons curve, a 90 dB intensity at 3 kHz results in the same feeling as 110 dB at 31 Hz (110-90=30 dB difference). The availability of a filter with a frequency response equal to the isophonic ure ne curves of Fig. 11.9 would allow an output similar to that of the human ear to be obtained and to devise an instrument capable of quantifying the perceived intensity. Taking into account that such a filter may not be straightforward to implement, simpler filters have been devised for this purpose. Fig. 11.10 shows > frequency weighting curv of filters that enable the intensity perceived by the human car to be quantified. The different curves of Fig. 11.10 take into account that the sensitivity is not the same for different sound intensities. Curve A is relative to intensities in the range 40 and 70 dB, curve B from 70 and 100 dB, curve C from 100 dB and higher. The sone is an alternative measurement unit for perceived intensity which some of the representing the frequency response was proposed by S. Smith Stevens so that 1 sone is equivalent to 40 phons. The sone scale is devised in a way that doubling the sone, the perceived intensity also doubles which corresponds to increasing the sound pressure level by a factor of 10 dB (ref. to ISO 226 [2003] and DIN 4563 / ISO 532 [1975] standards, based on the research work of E. Zwicker and H. Fastl). The suffix G (soneG) indicates that the loudne omputed from frequency groups, suffixes D or P refer to free field measurements (D and F stand for dir or free field), suffix R refer to room field or diffuse field. s in sone and phon of Fig. 11.11 shows that up to 30-40 phon the loudness in sone is almost constant. Above 40 phon the loudness increases rapidly. The relation between the loudne 11.2 Sources of Noise and Vibration 251 140 mob ! Y | 100[- ae 7 En sop An L] 2 2 sob y 5 f kHz Fig. 11.9. Isophonic curves: each curve indicates the intensity measured as a function of the frequency corresponding to the same perception of intensity heard. 11.2 Sources of Noise and Vibration The aim of the following section is to provide a short description of the main dynamic excitations that act on the car body: + road surface, + rolling of the tyre, + engine and powertrain, + brakes, + acrodynami 11.2.1 Road Surface The road profile is usually rather complex, potentially involves concentrated obs h as: bumps, depressions, potholes, tracks, junctions, distributed regular obs or random surfac asphalt, con- crete and dirt road. The main frequency content of these excitations is usually below 2002300 Hz depending on vehicle speed. As opposed to other sources, for a given speed, road excitations cannot be modified since they depend on ch a 254 11 Noise, Vibration, Harshness Tf all averages are not function of the coordinate x, the random process (the profile) is called stationary. In this case the explicit dependence on coordinate x can be omitted from the expression of the averages, which therefore can be indicated with reference to just the process variable (2): E[2], E[=?), R(Y). The averages defined in this way (Eq.s 11.11, 11.12, 11.13) consider, for a given coordinate x, the values corresponding to different profiles belonging to the same population formed by a large number of profiles. Thus they are considered to be ensemble average Conversely, it is also possible to take just one profile 2; analyze it as a function of the coordinate a: into account, and 1 (2i) = ML 2 (11.15) Í a (7) = lim E a). (11.16) Í Similarly, the autocorrelation function becomes: $) = (2 Jrilaj +A). (11.17) If the coordinate x can be considered to be a continuous variable over a travel distance £L, the values can be added with an integral instead of a sum of discrete values: L/2 () = Jim az), > ., (11.18) 1 L/2 ml, )dz, (11.19) 1 22 (a+) = lim > AA) = (zila) mf, zi + Ade. (11.20) A stationary process is said to be ergodic if all ensemble averages (those made considering for a given value of z different profiles belonging to the set of different profiles 2; ¿= 1,...,n) are the same as the corresponding averages along each profile (24): El2] = (2), (11.21) El] = (2), (11.22) RA) = 90). (11.23) Returning to the example of the driver that travels the same distance along different roads, if each road is of the same type, in a similar state of maintenance 11.2 Sources of Noise and Vibration 255 Fig. 11.12. Four different road profiles 2(x) as function of the traveled distance x on roads of the same type. and with a surface of the same type, it is reasonable to assume that the averages along a single path with varying x coordinate are the same as those on different paths (=;) at a given distance x, ¡.e. the random profile is ergodic. The autocorrelation function of an ergodic profile is be obtained by averaging the product of the profile z at two points, one leading the other by a fixed distance A (in [m)) L/2 1 PO a zu + Adu. (11.24) Instead of the distance A, the autocorrelation function can be expressed as func- tion of the number of eycles per unit distance y by means of its Fourier transform: R(A) = F S(v)e Adv. (11.25) Function S(v) is the power spectral density of the profile. If the signal z were a function of the time t, the autocorrelation function would be related to the time shift 7 (in [s]) betwes sliding along the data record, and the power spectral density to a time domain frequency (in [rad/s] or in [Hz]). From the de mensional point of view, the product vA in the complex exponential of Eq. 11.25 must be adimensional or, bearing in mind that eWA = cos(vA) + ¿sin(vA), should 256 11 Noise, Vibration, Harshness S(u) vV VHdV v Fig. 11.13. Power spectral density as function of the spatial frequency. The area under the graph is the root mean square of the profile. be expressed in [rad]. As A is a distance (in [m]), y is an angle per unit distance ([rad/m]), representing the inverse of a wavelength ([m/rad] or [m/eycle)). Apart from the factor 27, the power spectral density can be obtained as the direct Fourier's transform of the autocorrelation function S(w) = > Parera. (11.26) It can be demonstrated that S(») is an even function of the in the interval [—00, oo). Considering 11.24, for A =0 the autocorrelation function is equal to the mean square value (R(A = 0) = E[2?)). Substituting in Eq. 11.25, atial frequency », Ele R(0) = F S(v)d». (11.27) The result is that the mean square value is the area under the diagram of the power spectral density S() as function of the spatial frequency y (see, for ex- ample, Fig. 11.13). In other words, S(v)dv is the infinitesimal contribution of the frequency interval [v, v + dv] to the mean square value. As the total force FF acting on a beam due to a distributed load is the integral of the load per unit length p(x), similarly the power spectral density can be thought of as the density of the mean square value S(») per unit frequency ». From the dimensional point of view, the mean square value E[22] is a squared length ([m?), in SI units), the spatial frequency y is an angle per unit length ( [rad/mm]), and the power spectral density S(w) is therefore a squared length per unit spatial frequency ((m?/(rad/m)]=[m* /rad)). The definition of the power spectral density of Eq. 11.26 implies that the spatial frequency has positive and negative values. From both the experimental al point of view, the spatial frequency can have only positive val- s then usual to introduce an experimental power density Gexp(Vexp) [m?/(cycle/m)]. Gexp(Vexp) and S(v) [m?/(rad /m)] are related to each other, Gexp(Vexp) = 4715 (V). (1138) 11.2 Sources of Noise and Vibration 259 == => <=» <= == = LL 27 == == ÓN TE 1-2, E EN DT = 27h s gain 0 Spatial frequency [eycles/I] Fig. 11.16. The up and down rigid body mode of the car body is excited when the whcelbase includes an even number of hal£wavelengths. The pitching mode, by con- verse, is excited when the wheelbase includes an odd number of hal£-wavelengths. Road unevenness does not only affect the longitudinal direction but also the transversal direction, such that the right wheels move on a profile with similar but not identical characteristics (in terms of power spectral density, for example) as the left wheels. If the left and right wheels are excited out of phase there will be an excitation of the roll mode. The experimental measurements show that the road profile in longitudinal and transversal directions are not unrelated with each other. Irregularitics that develop in the longitudinal direction with a wavelength comparable to the wheelbase also develop in the lateral direction and affect all its track. In other words, the longitudinal and transversal wavelengths are similar and therefore long (compared to the wheelbase) wavelengths do not affect the roll motions. Conversely for short wavelengths the right and left profiles are little related with each other and may affect the roll motion. The road profile has been characterized as function of the longitudinal distance zx: from an arbitrary reference point leading to a power spectral density expressed as function of the spatial frequency. Dimensionally speaking the spatial frequency m] or [rad/1m], and the power spectral density of the profile in [m?/( This is because the road profile is a topographic variable and can be expressed directly as a function of geometrical coordinates on the road surface. However this enables the characterization of the road while revealing very little about the way the vertical dynamics of the vehicle would be excited by the road profile since the response on the vehicle is function of the time rather than the spatial dependence of the road irregularities. The link between these two depende is the speed V. 260 11 Noise, Vibration, Harshness It is then necessary to describe the road profile as function of the time fre- :ycles/m]) instead of the spatial frequency ([eycles/m)). B o s, the power spectral density must be expressed in [m?/(Hz)] instead of [m2/(cycles/m)]. Tf the vehicle travels with speed V, the contact patches of the wheels will move with frequency: f=Vv. (11.31) Eq. 11.31 can be substituted in Eq. 11.30: a) =0 (07 (11.32) the above substitution does not modify the power spectral density, that still rep- resents the density of the mean square value as function of the spatial frequency -=f G(f)dw, (11.33) o taking Eq.11.31 dv = df /V Ele] [Ps Pena the power spectral density G* is now consistent with the time frequency, ey 0 = Co NIG N, (11.34) Because coefficient N > 2, the power spectral density G* increases with the speed V. Fig. 11.17, shows the same power spectral density of Fig. 11.15 as function of the time frequency. The lower curve is relative to a speed of 50 km/h, the upper one to 150 km/h. For a given frequency the excitation increa sasing the speed. ses with incr Long wavelength humps and depressions A long undulation is a relatively smooth altimetric variation with a longitudi- nal dimension that is much larger than the occasionally by a depression of the road surfa this type of irregularity imply ess . In a similar way to the random profile, the frequency range involved by this excitation is function of the vehicle sp Fig. 1L1S shows the amplitude of the frequency spectrum of an undulation for different traveling speeds. The figure shows that the larger the speed, the larger the range of frequencies involved by the excitation, but the smaller the contributions at low frequency. Intuitive the time required to negotiate the obstacle redu ation frequency contact patch of the tires, caused for larger speeds and, correspondingly, the ez 11.2 Sources of Noise and Vibration 261 10* G(v) [m/Hz] q00 0" yor e ] Fina Fig. 11.17. Power spectral density of the road profile as function of the time frequency. a) 20 km/h, b) 180 km/h. increases. At 120 km/h a 6 m long, 30 mm de up to 10 Hz. As a rule of thum hump or depression h is L/V. The inverse of this time is approximately half the maximum ex frequency fmax ep depression involves frequenc f the vehicle moves with speed V and the Ss a length L, the time needed to negotiate the obstacle Ímax = aL. (11.35) Sharp obstacles Asperi covers are sharp variations of the road profile su ails :h as potholes, manhole . with wavelengths which are similar to or smaller than the lon- gitudinal size of the contact patch of each tire. The forces that the tire produces when negotiating obstacles of this type include significant longitudinal and vertical components. Additionally, if the ob- s oblique relative to the traveling direction, lateral components arise. Fig. 11.19 shows the variation of the vertical and longitudinal force the wheel hub as function of the traveled dis rectangular cross section. The t very low s on the rolling surface of the drum parallel to the axis. The distance between the wheel and drum axes is kept constant so as to provide an adequate preload ance when crossing a cle s made by rolling the tire on a large diameter steel drum at the ion is mounted eed of 2 m/s. Á steel beam of rectangular cr 264 11 Noise, Vibration, Harshness emi E zoo - z 3 g£ a Eno y : t : t t a] 00 1 Distance [mm] Fig. 11.21. Radial deformation of the tire obtained from the forces measured during the experiment of Fig, 11.19. Tire deformation (solid line) and profile of the obstacle (dashed) (height 25 mm, length 100 mm, 2 m/s speed). Similarly to long wavelength obstacle obstacl , the dynamic excitation due to sharp os is related to the vehicle speed. Fig. 11.22 shows the frequency spectrum of the equivalent obstacle of Eq. 11.37 at different speeds. As the size of the cleat (25 mm x 100 mm) is smaller than the depression of Fig. 11.18 (30 mm x 6 m), the time involved in crossing the cleat is shorter, at same speed, the excitation frequencies involved in crossing the cleat are larger. 11.2.2 Wheel The wheel subs; isc brake rotor, bearings) gives rise to various types of dynamic excitations due to rotation. The frequency of such excitations is related to the rotating speed 2. Mass unbalance Regardless of the nominal symmetry of the wheel, the center of gravity of the rotating part is never on the axis of rotation. The distance between the centre ontricity (6). The inertia forces due to rotation of the mass m of the wheel with eccez of gravity and axis of rotation is called ec € rotate with the same speed of the wheel on a plane perpendicular to rotation axis. The components of the unbalance force in an inertial frame x, 2 are: F, =m«2 cost, F.=meSPsin Qt; (11.38) the product me is the static unbalance. 11.2 Sources of Noise and Vibration 265 Amplitude [ram] 00 20 400 $00 200 1000 Frequency [Hz] 00 Fig. 11.22. Frequency spectrum of the equivalent obstacle. In a similar way to the centre of gravity, the main axis of inertia of the rotating part is never aligned with the axis of rotation: The angle (x) between these two axes is the dynamic unbalance, and the result is a torque about the x and 2 axes: My =-=x(h Ip) sin St, Me=x(J— Jp)52 cos Qt, (11.39) where J, is the polar moment of inertia about the wheel axis and and J, is the moment of inertia about an axis that lies on a plane perpendicular to the wheel axis and passing through the center of gravity of the wheel. The product x(J; — Jp) is the dynamic unbalance. Also in this case the exci- tation frequency is equal to the angular speed of rotation of the wheel. The static and dynamic unbalances can induce vibrations that can involve the steering wheel and, hence, can be perceived by the driver directly. 1f the exck tation frequency is the same as one of the natural frequencies of the suspension or the steering system, the amplitude of vibration can be amplified considerably by resonances, negatively affecting the level of comfort. The aim of the balancing operation is to reduce the dynamic forces acting on the wheel. Even if this does not modify the natural frequencies of the wheel and suspension, the effect is to reduce the amplitude of the vibrations, especially those of the steering wheel. Additional advantages are also a reduction of the tire tread wear, a reduction of the stresses on the bearings sion, and steering system, with an improvement of the fatigue life and re As on any other rotor, the aim of the balancing operation performed on the s to reduce the static and dynamic unbalance. This is done by adding (or eliminating) small balancing masses (m4) to a radius RR. (balancing radius), usually close to the outer diameter of the rim. If the unbalance added by the balanci s is of the same in magnitude and of opposite phase than the ving ma: original one, me, (11.40) 266 11 Noise, Vibration, Harshness the static unbalance is nulled. In other words: the effect of the balancing mas is to reduce (or eliminate) the distance between the center of gravity and the rotation axis. In practice the balancing operation will never be perfect, a certain amount of residual unbalance will always remain, depending on the accuracy of the instrumentation and of the procedure. To reduce the static unbalance of automotive wheels, it is possible to either add just one mass on the mean plane of the wheel, or add two masses (of half the size) at the same angular location but on opposite sides of the rim (on the rim flanges, for example) to compensate the static unbalance (me) without affecting the principal inertia axis of the wheel (angle x), while avoiding introducing a dynamic unbalance. Conversely, to reduce the dynamic unbalance, it is necessary to add two bal- ancing ma ill on opposite sides of the rim, but on opposite diametral loca- tions (180 deg) to reduce the dynamic unbalance (x) without affecting the static unbalance. Excitations due to tire shape and stiffness irregularities Because of the unavoidable tolerances in the shape and size of the tire and rim, the surface of the wheel will never be completely axisymmetric. The unavoidable errors with respect to an ideal toroidal surface can be expre harmonics of the angular coordinate about the wheel axis. Fig. 11.23 first four contributions of such series. The first harmonic ves an eccentrici a), the second an oval shape b), the higher order harmonics induce lobed « such as c) and d). Fig. 11.23 e) on the other hand, shows the basic mechanism that lea geometrical errors to the vertical and longitudinal forc shows just the first harmonic (eccentricity). As the radial stiffne much higher that of the suspension, the rolling radius is not constant even when traveling on a flat surface. The wheel axis is then subject to vertical displacements that induce vertical fore pension. Starting from relatively low vehicle speeds, the angular speed of the wheel is higher than the first natural frequency of the corner (Fig. 11.24) but lower than that of the unsprung mass. ads from the the figure of the tire is on the Es N< Lp. M. < Ma (11.41) This lar: implies that the displacements of the sprung mass induced by the irregu- s of the wheel are negligible compared to those of the unsprung mass. The vertical displacements of the wheel induce a movement of the suspension and, therefore, tra: :prung mass. In the frequency range between the natural frequencios of the sprung and the (Eq. 11.41), the amplitude of the radial excitation coming from smits forces to the UNSPIUNg Mas 11.2 Sources of Noise and Vibration 269 V [km/h] Fig. 11.25. Harmonic components of the tire vertical force F.. The peaks correspond to the excitation of the first resonant frequency (90 Hz) by different harmonics. larger contribution is due to the engine suspension vibrations (usually at about 15 Hz) and to the suspensions (1+2 Hz). When the frequency of one of the harmonic excitations due to the wheels is the same as the natural frequency of le body or one of its subsystems (for example the steering line, one of the vehicle body modes, the local modes of the panels, the acoustic cavity mod the response is amplified and induces a resonar the vel Tire tread The harmonic excitation due to the geometrical and stiffnes the tire is related to its structure and production tolera in the case of sleck tires. Another relevant source of dynamic excitation tread-road interaction, that involves mechanical and fluid dynamic effec road roughness, and the impact between the tread elements and the road surface, induce vibrations in the tire carca by the friction forces exchanged through the contact patch. Because of the tire s, these forces increase from the leading edge of the contact patch to the Qualitatively speaking, each part of the tread entering the contact is not deformed in the tangential direction. Because of the different speed between the ground and the ca the tangential deformation increases and reaches its > to the trailing edge of the contact patch. is amplified at high temperature. In this case the tread pattern 'k to the road surface (stick snap) due to the adhesion forces, and the accumulated tangential deformation is released suddenly when the contact ends causing tangential vibrations. As in the previous cases, the excitation frequency related to the contact of the tread pattern is related to the angular speed of the wheel. At 120 km/h maximum This effes elements 270 11 Noise, Vibration, Harshness 0 20 40 60 80 100 S (Mz Fig. 11.26. Frequency spectrum on a seat rail at 120 km/h. In evidence the harmonics of the wheels angular speed. Fig. 11.27. Dynamic excitations due to the tread. a) radial and tangential resonance of the tread elements; b) cavity resonance of the air trapped in the tread pattern. the fundamental excitation of a tread pattern with 60 elements of the same size is equal to about 1.2 kHz. The adoption of a tread pattern with a size that varies randomly along the tire circumference allows to reduce a pure tone excitation frequency when traveling. The effect of the random variation of the tread pattern is to make the impact of following elements of the pattern not periodic. The periodicity is thus that of the wheel rotation, with reduced high frequency vibrations that fall in the acoustic range. The vibrations induced by the tread pattern are transmitted to the tire that related to the carcass and air are transmitted to the rest of the vehicle through vibrates with its own dynami cavity within. These vibrations the structure or to the air surrounding the wheel. Fig. 11.28 shows the result of a noise intensity measured at the driver's left ks are due to the resonance of the tire structure (A), and ear position. The pe: to the internal air cavity (B). The frequency of these peaks is constant; the third peak (C) is proportional to the vehicle speed since it is caused by the excitation due to the tread pattern. 11.2 Sources of Noise and Vibration 271 a 2 z Q ee ¿EY YB822382 100 160 250 S00 1000 2000 4000 10000 Frequency [Hz] Fig. 11.28. Noise intensity at the left driver's ear. The measurement is performed on a drum test bench at 120 km/h. A) structural resonance of the tire, B) internal air cavity resonance; C) tread contact excitation. 1) left and right smooth drum surface; 3) left and right rough drum surface; 2) left drum with rough surface, right drum with rough surface. Aerodynamic phenomena caused by the wheel close to the contact patch con- of noise. The so-called air pumping relates to the reduction of the volume produced by the tire deformations at the leading edge of the contact patch followed by an expansion at the trailing edge. The variations of the volume produce an air flow in the channels between the tread pattern grooves and the ground. The complexity of the grooves induce vibrations in the air with an amplification of the vibrations through an organ pipe mechanism. stitute another relevant sou 11.2.3 Engine The internal combustion engine produces vibrations due to the movement of its mechanical parts and to the thermodynamic eycle. The crankshaft and the alternating m of the pistons and connecting rods generate centrifugal and alternate inertia forces that are transmitted to the engine block. The inertia forces can be computed to a first approximation by considering the masses of the piston (mp), and the crankshaft (ma), while the contribution of the connecting rod (mp) can be split in two parts, added to the piston (my/3) and to the crankshaft (2/3 mp). 274 11 Noise, Vibration, Harshness » La : 2” En E” Es E Él 7 : d Rh (q——_————_——— Cycle position [deg] Fig. 11.30. a) Pressure as function of the crankshaft angle; b) torque on the crankshaft. the resonance of one of these modes (in this case the engine block) excited by the various engine orders. The vibrations of the engine block surfaces also induce pre the engine compartment that propagate as air borne noise, transmitted to the interior of the car through the vibration of the structural panels (the firewall, for example). Fig. 11.32 repre the contribution of the different parts of an engine to the total emitted acoustic power. In addition to the combt , the intake and exhaust flows constitute of dynami tation. The pre gas flow, that induce vibrations in the intake and exhaust pipes that radiate then noise in the engine compartment. Some of the vibrations are also transmitted to the engine block that propagates it and radiates other acous Fig. 11.33 shows the frequency analysis close to the output of the exhaust pipe during a ramp up. The part related to the engine orders is due to the combustion and to the periodic motion of the gas flow through the exhaust A considerable part of the spectrum is constituted by wide band and uniform excitation due to the turbulence and other aeros phenomena. Apart from the noise radiated by its surface, the low frequency vibrations of the exhaust pipe are also transmitted directly to the vehicle body by structure borne paths through the suspension elements used to connect it to the vehicle underbody. The fuel injection compre tioning system, and the power steering system are some of the various aux powered by the engine. The different mechanisms involved in their opera (mechanical, fuid dynamics, electrom 1) correspond to other soure and vibration, that are transmitted by air borne paths or by the strt The pumps also transmit noise through pres in the pipes. Usually the accessories are powered by the belt, chain or other type of drives with a ¡on ratio typical of each auxiliary which also influences the presence jure waves in ents another sou ure variations in the turbulent valves. stem, alternator, fan, starter. sor of the air condi- amic; Nois: sure waves transmi 11.2 Sources of Noise and Vibration 275 Fig. 11.31. Frequency spectrum of the acceleration on the engine block. 8 cylinders engine. For each engine speed (rpm), the amplitude of the harmonics is reported with a color coding (grey-scale) as function of the frequency. The dotted line evidences the fourth harmonic. of principal harmonic components of the noise and vibration as function of the engine speed. Chain and timing belt drives are the two main solutions used to drive the camshafts, both of which produce noise during operation. Roller chains produce a noise related to the impact of each roller against the toothed sprocket and the slipping against other parts such as tensioners and guiding elements. Timing belts produce noise because of the contact of each tooth against the pulley. The high internal damping of the rubber that constitutes the belt reduces all harmonic components except the fundamental to negligible levels. A possible source of noise in this case is the pumping of the air to and from the volume that could be trapped between pulley and belt. At high temperatures, as the surface of the belt becomes stickier, the belt drive may emit wide band noise due to the stick and slip between the belt and the pulle s is especially relevant for non synchronous belts used in auxiliaries s. Apart from the contact against the pulleys, belt transmi can produce noise because of the lateral vibrations of the segments traveling from one pulley to the next. These vibrations can be excited at resonance by the engine orders and the geometrical defects of the pulleys. Apart from the noise ns 276 11 Noise, Vibration, Harshness Fig. 11.32. Cumulative contributions of the total acoustic power emitted by the engine. 1) high pressure pipes, 2) fuel pump, 3) engine head, 4) exhaust, 5) engine block, 6) intake, 7) head covers, 8) oil sump. and vibration issue, such vibrations can induce fatigue in the drive that affects its reliability. A completely different excitation is produced by the power train because of sudden variations of the throttle command. This happens when the driver pushes on or releases very fast the accelerator. Similarly, the sudden engage- ment/disengagement of the clutch produce impulsive torque excitations to the power train. The result is a torsional vibration of the transm vibrations are transformed by the wheels in longitudinal vibrations of the whole vehicle that could involve frequencies up to about 10 Hz (Fig. 11.34). ion. The torsional 11.2.4 Transmission The gearbox is a source of noise and vibration due to the meshing between the gears. The most important sources of gear noise are rattle and whine. In most of the gearboxes for automotive applications all gear sets are con stantly meshing together; at any specific moment, while the engaged set is trans mitting power, the others run idle. Because of the torque fluctuations, the shafts are subject to torsional vibrations, this induces idle gears to bounce against each umferential gap in between. Even though small, this gap is y for the functionality of the gear set to allow lubrication of the teeth during contact and avoid them seizing. The rattle can be amplified by the torsional resonances of the transmission (usually below 100 Hz). The acceleration measured on the gearbox during rattling is shown in Fig. 11.35. The impulses may be not always repeat period; and can change with time and depending on the nature of the excitation coming from the other elements of the transmission. The gear whine, is the noise produced at the harmonics of the tooth passing frequency. It is generated by the transmission error at the loaded gear meshes other in the es 11.2 Sources of Noise and Vibration 279 Lo T = "Impact Period" Acceleration 1 1 1 0.00 0.05 0.10 0.15 0.20 0.25 0.30 time [s] Fig. 11.35. Acceleration measured on the gearbox during rattling of a gear set. The impulses correspond to the bouncing of the gears against each other in the (small) circumferential gap between them. oting each span, the bow of their axis, and the non y of the supports induce dynamic exc the shaft is rotating. These excitations are transmitted to the vehicle structure via the supports. The unbalance ex ations when itation is amplified when the angular speed of the shaft is the same as one of its natural frequencies (see [25)). Universal joints are sometimes installed along prop shafts to compensate for the relative displacement and misalignment between the two ends. Considering a single universal joint, the transmission ratio between the input and output speeds (w; and ws) is not constant Wo cos 8 2 - —_— 11.48 Wi sin? Ba ( ) where (3 is the angle between the input and output shafts and ó; is the rotation of the input shaft. Because of the variation of the transm transmitted torque is not constant. Under the assumption that the transmis ipate power it must be ¡on ratio also the ion Mw. = Mw; (11.49) the output torque M, is then (11.50) Tf the input torque M; is constant, the output one has a ripple between the two values (11.51) 280 11 Noise, Vibration, Harshn 133. Hz Spectrun 166 Hz Topo E O rroor=os: Ss -] Las, -Q ¿ +:cot-0s L—| fB=500/60*28*3=700 Hz. sl i : Mola de E 1.00£-07 HR 200 1060 1500 7006, f [Hz] Fig. 11.36. Frequency spectrum of the transmission error of a 28 teeth gear running at 500 rpm. The peaks correspond to the fundamental meshing frequency and its har- monics ment; . input and output shaft torques, M, constraint torque due to supports. b) torque diagram corresponding to case a). c). Shaft with two universal joints that compensate the transmission error. 11.2 Sources of Noise and Vibration 281 Fig. 11.37 b) shows the free body diagram of the universal joint of Fig. 11.37 a). M, is the reaction torque of the bearing support. As the output torque M, is variable, also the bearing reaction torque M, is variable and can potentially induce vibrations of the support and the neighboring structure. Additionally, the angular accelerations of the output shaft require inertia torque components: sin? Bcos Bsin 26; : (11.52) * (1 - sin? Bos? 64)" Wo due to the not negligible inertia of the output shaft. The term sin 26; at the numerator of Eq. 11.52 indi harmonic of the inertia torque acting on the output shaft of Fig. 11.37 is twice the angular speed of the input shaft. From the purely kinematic point of view, a shaft including two universal joints with the same orientation and same misalignment $, allow a constant transmis sion ratio between the first and the last shaft (Fig. 11.37 c) to be obtained. From the qualitative point of view, the speed variations that are induced in the intermediate shaft by the first joint are compensated by the second. In any the intermediate shaft is subject to angular accelerations that induce variable inertia torques, adding variable components to the supports in sed by the variable torque (M,) transmitted to the inter- s that the fundamental addition to those ca mediate shaft. The variability of the loads acting on the prop shaft support bearin high levels of vibration in the vehicle structure. Therefore elastomeric elements are usually integrated to filter such vibrations. s can cause 11.2.5 Brakes The noise and vibration phenomena that may be involved during braking are many, as the terms used to indicate them. At low frequency (010 Hz) the ABS system generates pre hydraulic circuit that are transmitted to the driver through the brake pedal. The variations of the braking torque determine a variation of the tire brake force that leads to longitudinal vibrations of the suspensions and the vehicle, The vibrations at frequencies between 3040 Hz and 100 Hz are indicated as judder or shudder and can be perceived on the steering wheel, the brake pedal and the floor: Such vibrations are due to a ripple of the braking torque about a mean value that can be caused by a number of factors, for example: a not flat surface of the d stic instabilities in the brake, uneven deformations of the brake during its installation, etc... The typical feature of the judder is the dependency of its frequency from the angular speed of the wheel, confirming that the primary cause is related to the unevenness of the friction surface. When the temperature of the brake is high 'ed by the weld- ing on the disc of some material of the pads, altering the surface properties and causing a torque ripple. Similarly, an irregular surface of the disc can cause hot re waves in the hermo-e! judder can be 284 11 Noise, Vibration, Harshness Fig. 11.39. Aerodynamic vortexes around the vehicle body. The diagram shows that for increasing speeds the contribution of the aerody- namic noise incre: more rapidly than the other contributions, until it becomes dominant at speeds typical of traveling on a highw 11.3 Dynamic Behavior of the Body and Modal Analysis 11.3.1 Dynamic Equations Under the assumption of the small displacements, the velocity vector of a generic point of the structure can be expressed in an inertial reference frame as: v=aj+bq+c, (11.56) where q = (41,42,..-,4)" is the column matrix including all Lagrangian co- ordinates of the em. Such coordinates enable the deformed configuration of the structure to be determined as function of time t. If the displacements are negligible compared to the of the position of the point (a From the dimensional point of view, c is a speed column matrix, the elements of se of the structure, matrices a, b, c are a function 2) but not of time £ nor the displacements q. matrix b are the inverse of a time and the elements of a are number In the case of a continuor described by an infínite number of Lagrangian coordinates g denoting the displacement of each structure, the deformed shape i 11.3 Dynamic Behavior of the Car Body and Modal Analysis 285 Fig. 11.40. Vortex shedding in the wake of a long cylindrical body in a flow perpen- dicular to its axis. The structure of the wake is characterized by vortexes that separate from the cylinder on opposite sides of the cylinder section. This wake is usually re- ferred to as von Karman wake. The vortex that is going to leave the cylinder generates a downward force. the Lagrangian coordinate is a y, 2,t). ontinuous function g(«, y, 2,t) ¡ption of the dynamic behavior in point of the continuous material. In this c continuous function of the position and time g(w Considering the Lagrangian coordinates of the position and time would lead to a dex elements method is the ability to discretize the structure: ¡.e. to approximate Eq. 11.56 with a finite number of displ ingle element represents a limited portion of the structure the displ and speeds of which are approximated by a finite number of Lagrangian coordinat These coordinates, called element degrees of freedom, represent the displa 1 points ngle element, matri and e can be determined from the so-called shape functions by taking the of the element called nodes. Considering a s speci a, kinematics of the element into account. The expression of the speed v allows to obtain the kinetic energy, that includes the contribution of each elemental volume of mass dm. 1 T= 3/ v dm = 2 vol 23/ (Ctra (1157) vol + 707 bq + 247 ac +27 be + Tc) dm 286 11 Noise, Vibration, Harshness 80 70 » < 60 8 ¡50 540 2 » ANS A 20 TeeEsessssssss 988528928888 838rm Fig. 11.41. Spectrum of the aerodynamic noise at the driver's left ear location during aeroacoustic wind tunnel tests. 1) standard vehicle; 2) taped doors; 3) no mirrors no wipers. 97 ss so a Za) roal 3 7] “Sy == 5 65 so << Asroacostic tunnel ss so + 45 40 20 40 60 so 100 no 140. 160 180 V [km/h] Fig. 11.42. SPL measured on road and wind tunnel tests. The aerodynamic noise increases with speed.