Modelling of Pneumatic Engine Mount Master’s Thesis in the Master’s programme in Sound and Vibration AMÉLIE RENAULT Department of Civil and Environmental Engineering Division of Division of Applied Acoustics Vibroacoustics Group CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2005 Master’s Thesis 2005:48 MASTER’S THESIS 2005:48 Modelling of Pneumatic Engine Mount AMÉLIE RENAULT Supervisors: Lars H. Ivarsson (Department of Applied Acoustics) Arndt Graeve (Trelleborg Automotive) Department of Civil and Environmental Engineering Division of Applied Acoustics Vibroacoustics Group CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2005 TRELLEBORG Automotive Modelling of Pneumatic Engine Mount  AMÉLIE RENAULT, 2005 Master’s Thesis 2005:48 Department of Civil and Environmental Engineering Division of Applied Acoustics Vibroacoustics Group Chalmers University of Technology SE-41296 Göteborg Sweden Telephone +46 (0)31-772 1000 Reproservice / Department of Civil and Environmental Engineering Göteborg, Sweden 2005 Modelling of Pneumatic Engine Mount AMÉLIE RENAULT Department of Civil and Environmental Engineering Division of Applied Acoustics Vibroacoustics Group Chalmers University of Technology Abstract The goal of this Master thesis has been to build a mathematical model of a pneumatic damped engine mount, which could be a cost effective alternative to hydraulically damped engine mount. This model will help to understand the pneumatic mount and its damping mechanism. An engine mount must satisfy at least two essential but conflicting criteria. First, the mount has to carry the static gravity load of the engine. The second requirement is that the compartment should be isolated from vibrations and acoustically comfortable: the mount has to be compliant and lightly damped for small excitation amplitudes and over the higher frequency range. The third requirement is to isolate the car from the motions due to bumpy roads, abrupt vehicle acceleration or deceleration, and braking and cornering. That means that the mount has to be stiff and highly damped Conventional rubber mount cannot satisfy both requirements simultaneously and a compromise between resonance control and isolation is inevitably needed, which is to get non- linear stiffness and damping dependant on frequency. There are many disadvantages to the hydro mounts: it is big, heavy, complicated and expensive. Then, an air damped engine mount could be an issue to decrease the size of the damping device (it has only one chamber) and lower the price (it is quite simple and is based on air compression). But a pneumatic mount is less damped than the hydro mount and the damping has to be increased in order to damp heavy engines. In this report will be presented an overview about hydro-mechanical mount, pneumatic mounts and some tools that will be used to model the rubber dynamic properties, the air volume and the orifice behavior. After that, the model will be explained. Finally, some proposals how to increase the damping will be explained. After having understood how an hydraulic mount is working and explored some simple pneumatic spring, the model of the pneumatic mount has been implemented in MATLAB and finally, the calculated curves are compared to the measured one. The mount has physically been divided in three parts in order to model it more easily and the conclusion of the model- experiment study shows that: • The rubber part is well modelled with a spring in parallel with a dashpot, • The orifice impedances seems to be quite correctly modelled since the global model predicts really well the resonance frequency • The air compliance is not modelled accurately enough. It can be also remarked that the simplified orifice impedance is working well for the chosen frequency range and for this application. Keywords: Engine mount, pneumatic mount, damping modelling, oriface damping, engine suspension. 2 Acknowledgments It was really great to do my master thesis at the acoustic department. I would like to thank Lars Ivarsson for being a patient and good supervisor, Wolfgang for helping me with the physical understanding of my subject, Börje for fixing my different computer problems and of course also Gunilla for all administrative things. Thanks to all my friends for the great time, which allowed me to spend a less boring thesis. I will never forget the balcony discussions and the ‘fika’ time which allow to take energy and forget the stress... To my little sister who waited such a long time for me to come back home. 3 I. Introduction The goal of this Master thesis is to build a mathematical model of a pneumatic damped engine mount, which could be a cheap alternative to hydraulically damped engine mount. This model will help to understand the pneumatic mount and its damping mechanism. An engine mount must satisfy at least two essential but conflicting criteria. First, the mount has to carry the static gravity load of the engine. The second requirement is that the compartment should be isolated from vibrations and acoustically comfortable: the mount has to be compliant and lightly damped for small excitation amplitudes and over the higher frequency range. The third requirement is to isolate the car from the motions due to bumpy roads, abrupt vehicle acceleration or deceleration, and braking and cornering. That means that the mount has to be stiff and highly damped Conventional rubber mount cannot satisfy both requirements simultaneously as the lumped stiffness, kr, and the viscous damping coefficient, br, in the shear mode are nearly invariant with excitation amplitude and frequency over the concerned excitation range of vehicle systems. Thus, a compromise between resonance control and isolation is inevitably needed, which is to get non- linear stiffness and damping dependant on frequency. The hydro mount is typically optimized for placement, orientation, kr and br in tuning the inertia channel and decoupler, which means adapting the inertia channel and decoupler dimensions in order to get the required resonance frequencies. There are many disadvantages to the hydro mounts: • Since it has to carry the static load of the car weight it is big and heavy • In order to fulfill the requirements, the device is complicated (two chambers filled with a special liquid with a membrane and a channel separating them) and it is expensive. Then, an air damped engine mount could be an issue to decrease the size of the damping device (it has only one chamber) and lower the price (it is quite simple and is based on air compression). But a pneumatic mount is less damped than the hydro mount and the damping has to be increased in order to damp heavy engines. In this report will be presented an overview about hydro-mechanical mount, pneumatic mounts and some tools that will be used to model the rubber dynamic properties, the air volume and the orifice behavior. After that, the model will be explained. Finally, some proposals how to increase the damping will be explained. 4 I. INTRODUCTION......................................................................................................................................................... 3 II. THE HYDROMOUNT FUNCTIONING AND MODELLING .............................................................................. 5 1. SCHEMA...................................................................................................................................................................... 5 2. PARAMETERS .............................................................................................................................................................. 5 3. HYDROMOUNT FUNCTIONING ..................................................................................................................................... 6 4. FREQUENCY RESPONSE CHARACTERISTIC ................................................................................................................... 6 5. MODELLING OF AN HYDROMOUNT WITH DECOUPLER ................................................................................................. 7 III. PNEUMATIC SPRINGS: SOME EXAMPLES.................................................................................................... 11 1. USED PARAMETERS AND CONSTANTS........................................................................................................................ 11 2. DESCRIPTION OF A BASIC PNEUMATIC FLEXIBLE ELEMENT........................................................................................ 12 3. SINGLE ACTION PNEUMATIC FLEXIBLE ELEMENT WITH DAMPING CHAMBER ............................................................. 12 IV. MODELLING OF THE RUBBER ......................................................................................................................... 15 1. USED PARAMETERS................................................................................................................................................... 15 2. RUBBER YOUNG’S MODULUS AND SHEAR MODULUS ................................................................................................ 15 3. DEFINITION OF THE ELASTIC STIFFNESS .................................................................................................................... 16 4. DEFINITION OF DAMPING .......................................................................................................................................... 17 5. PRESENTATION OF DIFFERENT VISCOELASTIC MODELS ............................................................................................. 18 V. A MODEL OF THE ORIFICE IMPEDANCE....................................................................................................... 19 1. PARAMETERS USED AND IMPEDANCE DEFINITION ..................................................................................................... 19 2. INNER IMPEDANCE .................................................................................................................................................... 20 3. RADIATION IMPEDANCE ............................................................................................................................................ 22 VI. ELECTRO- MECHANICAL-ACOUSTICAL ANALOGY................................................................................. 23 VII. THE MEASUREMENTS....................................................................................................................................... 25 1. GENERAL SET-UP ...................................................................................................................................................... 25 2. STATIC MEASUREMENT ............................................................................................................................................. 26 3. DYNAMIC MEASUREMENTS AND RESULTS................................................................................................................. 27 4. PNEUMATIC ENGINE MOUNT WITH CLOSED ORIFICE .................................................................................................. 28 5. THE PNEUMATIC ENGINE MOUNT WITH OPENED ORIFICE ........................................................................................... 29 VIII. DESCRIPTION OF THE MODEL ..................................................................................................................... 30 1. UNKNOWNS AND ASSUMPTIONS................................................................................................................................ 30 2. THE AIR COMPLIANCE EXPRESSION ........................................................................................................................... 31 3. THE RUBBER MODEL USED ........................................................................................................................................ 31 4. THE BASIC EQUATIONS USED FOR THE FLUID............................................................................................................. 32 5. THE LIMITS OF THE MODEL........................................................................................................................................ 33 6. NUMERICAL IMPLEMENTATION IN MATLAB ........................................................................................................... 33 7. PARAMETER STUDY .................................................................................................................................................. 34 IX. MODEL/MEASUREMENTS COMPARISON ..................................................................................................... 36 1. THE RUBBER MEASUREMENT AND MODEL COMPARISON ........................................................................................... 36 2. MEASUREMENTS AND MODEL FIT.............................................................................................................................. 36 X. PROPOSALS HOW TO INCREASE THE DAMPING ........................................................................................ 38 1. SEVERAL HOLES, ORIFICE AREA INCREASE................................................................................................................ 38 2. ACOUSTIC ABSORBENT INSIDE THE VOLUME............................................................................................................. 39 XI. CONCLUSION AND DISCUSSION...................................................................................................................... 40 REFERENCES ............................................................................................................................................................... 41 XII. APPENDIX A1: PARAMETER STUDY FOR –850 N PRELOAD AND 0.1 MM AMPLITUDE.................. 42 XIII. APPENDIX A2: PARAMETER STUDY FOR –850 N PRELOAD AND 1 MM AMPLITUDE................... 46 XIV. APPENDIX B: CALCULATED AND MEASURED CURVES COMPARISON ........................................... 50 XV. APPENDIX C: MATLAB CODE.......................................................................................................................... 56 5 II. The hydromount functioning and modelling 1. Schema Figure 1: Schema of an hydromount 2. Parameters With the help of the article [1], the useful parameters can be listed: Parameters: Ad Ai Ar di d1, d2 F li lr Diaphragm area, [m2] Cross-sectional area of the inertia track, [m2] Equivalent piston area of the rubber, [m2] Hydraulic diameter, [m] Chambers diameters, [m] Applied force, [N] Effective length of the inertia track, [m] Thickness of the rubber up, [m] mE mr p The engine mass, [kg] Effective mass of the rubber in the shear mode, [g] Pressure, [Pa] qd ,qi Fluid flow rate through the decoupler and inertia track, [m3/s] x Displacement, [m] t2 Thickness of the rubber below, [m] V1, V2 Chambers volumes, [m3] Constants B Cd l Er Rd µ ν ρ Bulk modulus of the fluid, [Pa] Equivalent linear compliance, [m5/N] Young’s modulus of rubber, [Pa] Linear fluid resistance, [N/m5] Fluid viscosity, [Pa.s] Poisson ratio, [-] Fluid density, [kg/m3] Upper chamber Lower chamber Decoupler Inertia track 6 Then other parameters can be computed using the geometry characteristics and the different constants. Lumped fluid resistance 4 i i i d 128R π µ = l Track inertance i i i A I lρ = Fluid mass ρV1 and ρV2 Compliances r2 4 2 l 2 2221 d l 1 12 l 1 r 2 32 rr11 Et16 d B2 VCC C B2 VC B2 V E)1( 231AC π +≈= +≈ + υ− υ−υ− ≈ l Inertances of the chambers 2 r 2 2 r 1 A m, A m 3. Hydromount functioning In references [1] and [10], the functioning of a hydromount is described. The two chambers hydro- mechanical mount consists of two fluid chambers filled with glycol, which communicate through an orifice and an inertia track or a damping channel. In the upper chamber, a steel casing and an elastomeric spring (primary rubber) supports the static load and are supposed to act uniaxially through the mount axis. The primary rubber also serves as a piston to pump the fluid through the rest of the mount. The lower chamber, also made of rubber, is designed to be very soft and is used to absorb the transferred fluid without considerable pressure. Damping is created by the inertia effect of the fluid column resonating between the two springs. The inertia track is a lengthy spiral channel that enables the fluid to pass from the upper chamber to the lower chamber. The inertia effect increases the dynamic stiffness. In fact, at high frequencies, the fluid doesn’t follow the excitation and the orifice channel closes; the increasing pressure in the main chamber increases the stiffness. The decoupler is a plastic plate. For small amplitude excitation, most of the fluid transport between chambers is via the decoupler orifice. For larger amplitude excitations, the decoupler is blocked and the most of the fluid flow is forced through the inertia track. The geometry of the orifice channel (length, cross section), which determines the frequency of maximal angle and thus the damping properties, is used to tune the mount as a purpose of control. Some other important parameters determining the behavior of the entire mount are the hydraulic area and the volumetric stiffness. 4. Frequency response characteristic The frequency response characteristic is described by J. E. Colgate et al. in [1]. Here is a summary: • Below 5 Hz, the fluid flows between the chambers with low resistance. • From 5 to 14 Hz and amplitude from 0.5 to 5.0 mm, these excitations are on the range of engine resonance and large enough to require significant damping. The resonance frequency of the system is on this range. 7 • From 25 to 250 Hz and amplitude from 0.05 to 0.5 mm, these excitations can cause noise and vibration and require good isolation. The resistance force to the fluid flow through the opening is greater than the elastic resistance of the rubber spring. The fluid pressure in the upper chamber fluctuates with excitation frequency but fluid pressure in the lower chamber remains almost constant. 5. Modelling of an hydromount with decoupler In this part is explained how to draw a simple model of a hydro-mechanical mount with decoupler and how to get the parameters. It is also shown how to simplify a model for low frequency since a model for low frequencies has later to be derivate. R. Singh et al. developed in [2] a lumped parameter model of a hydro-mechanical mount. Only vertical motions are taken into account here. The first step is to write the equation of motion and the momentum balance for each volume and mass involved in the model. The rubber is here modelled with Voigt model. Ft is the total excitation force )t(FF)t(F s t += (4.1) xt is the total displacement )t(xx)t(x s t += (4.2) The indice s stands for static conditions: Fs is the static load corresponding to the engine weight and xs is the corresponding static displacement. F(t) and x(t) are the fluctuating components. kr and br are assumed to be invariant with ω and xt. The figure 2, which is a sketch of the forces acting on the mount, is related to the equations (4.3) to (4.10). Figure 2:Schematic of a hydromount with a free decoupler The equation of motion of the rubber mass can be written as: Ft (t) − krx t (t) − br Ý x t (t) − Ar p 11 t (t) − patm (t)( )= mrÝ Ý x t (t) (4.3) The momentum balance applied to the fluid mass gives the three equations (4.4) to (4.6). F patm p11 p22 p21 p12 qd qi 2 b r 2 k r 2 br 2 k r mr x x2 x1 8 For the fluid in the upper chamber ( ) )t(xmA)t(p)t(p t 1r tt 11211 &&=− (4.4) For the fluid in the lower chamber ( ) )t(xmA)t(p)t(p t 2r tt 22112 &&=− (4.5) For the inertia track volume and the decoupler volume: ( )tqRqI)t(qR)t(p)t(p t i t i t d tt iid2112 +==− & (4.6) The equation of continuity applied to the compliant volume chambers gives, for the chamber 1: ( ) )t(pC)t(x)t(xA 1111 tt r 1 && =− )t(pC)t(q)t(q)t(xA t 12 ttt r 12id1 && =−− (4.7) (4.8) Then for the chamber 2: )t(pC)t(xA t 22 t r 222 && = )t(pC)t(xA)t(q)t(q t 21 t r tt 212id && =−+ (4.9) (4.10) ,C,C,C,C 22211211 are the compliances. The two first are the upper chamber compliances and the two last are the lower chamber compliances. The static equilibrium gives: )pp(AxkF atms,11rsrs −+= , s22211211sr p)CCCC(xA +++= (4.11,12) and s222112s,1r p)CCC(xA ++= , s22s,2r pCxA = (4.13,14) In considering the hydro-mechanical mount with a fixed decoupler (inactive decoupler), which means qd(t)=0 and eliminating the pressure variables, one gets the following time-varying equations (4.15) to (4.18)with r i i A )t(q)t(x =& , and for the time varying part of the variables. ( ) ( ) ( ) ( ) ( ) ( ) ii2i 21 2 r i1 12 2 r i 2 ri 222 22 2 r 2i 21 2 r 11i1 12 2 r 1 11 2 r rrr1 11 2 r xm)t(x)t(x C A)t(x)t(x C A)t(xAR )t(xm)t(x C A)t(x)t(x C A )t(xm)t(x)t(x C A)t(x)t(x C A )t(xm)t(xk)t(xb)t(x)t(x C A)t(F &&& && && &&& =−−−+− =−− =−−− =−−−− (4.15) (4.16) (4.17) (4.18) 9 Then, the analogous mechanical system can be drawn as seen in figure 3. Figure 3: analogous mechanical system with free decoupler The parameters are defined as: 22 2 r 22 21 2 r 21 12 2 r 12 11 2 r 11 C Ak; C Ak; C Ak; C Ak ==== and i 2 rii 2 ri RAb;IAm == . The different k correspond to the chamber stiffness, the mass mr to the rubber mass, m1 and m2 to fluid mass in the chamber 1 and 2 respectively. bi corresponds to the inertia track resistance. Then, in [2] again, the equations are simplified for low frequency. In fact, the impedances of the lumped fluid masses are negligible at low frequencies and a two-degree-of-freedom model can be derived. Then, there will only be two compliances C1 and C2, which can be calculated as the sum C11+C12 and C21+C22 respectively. Moreover, )t(p)t(p)t(p 11211 =≈ and )t(p)t(p)t(p 22221 =≈ . )t(xm)t(pA)t(xb)t(xk)t(F r1rrr &&& =−−− )t(qR)t(qI)t(qR)t(p)t(p iiiidd21 +==− & )t(pC)t(q)t(q)t(xA 11id1r && =−− )t(pC)t(q)t(q 22id &=+ (4.19) (4.20) (4.21) (4.22) In eliminating the variables p1(t) and p2(t) and with r i i A )t(q)t(x =& one gets: ( ) )t(xm)t(xk)t(xb)t(x)t(x C A)t(F rrri 1 2 r &&& =−−−− ( ) )t(xm)t(x C A)t(x)t(x C A)t(xAR iii 2 2 r i 1 2 r i 2 ri &&& =−−+− (4.23) (4.24) Then, a simplified model for hydromount with fixed decoupler (which correspond to low frequency), can be drawn as in figure 4. m2 mi br mr bi k11 k12 k21 k22 m1 xt(t) x1 t(t) xi t(t) x2 t(t) kr Ft(t) 10 Figure 4: analogous mechanical system with fixed decoupler This was however a linear model and it is possible to use it because high and low frequencies can be separated and then get two linear models in order to cover the entire frequency range. But it is sometime required to derivate a non-linear model. A. Geisenberg et al., in [12], explain how to get a non-linear model in enhancing the parameters of the linear model discussed above. The parameters identification via experiment is also well described in this paper. mi kr br mr bi k1 k2 xt(t) xi t(t) Ft(t) 11 III. Pneumatic springs: some examples In this chapter will be described two simple pneumatic springs: a basic cylinder piston element and a single action pneumatic element with damping chamber. In both cases the stiffness will be derived. 1. Used parameters and constants Parameters A d F h l m T0 w Q V W ν Piston cross-sectional area, [m2] Diameter of the capillary, [m] External applied force, [N] Height (smaller dimension), [m] Length of the capillary, [m] Mass of preload, [kg] Initial temperature, [K] Width (larger dimension), [m] Volume flow rate, [m3/s] Instantaneous internal volume between piston and cylinder, [m3] Weight flow rate (W=Q.g), [N/s] Flow velocity, [m/s] Constants g patm R γ µ ρ Gravity acceleration, [m/s2] Atmospheric pressure, Patm=101300 Pa for the air Universal gas constant, R=286.4 N-m/(kg K) for the air Ratio of specific heats, γ=1.3 for the air Dynamic viscosity of the gas, [Pa.s] Gas density, [kg/m3] Computed parameters: Reynolds number Round capillary: µ ρνdRe = Rectangular capillary: µ ρ = µ ρ = w Qh wh QRe Capillary resistance Round capillary: lµ π = 128 dC 2 c Rectangular capillary: lµ = 12 whC 2 c Initial pressure A mgpp atmi += 12 2. Description of a basic pneumatic flexible element The book [7] describes among other systems of vibration isolation, pneumatic flexible element, which are using compressed air as flexible element. The more simple is a linear cylinder-piston system. The piston is moving into the cylinder, which can be of any cross-sectional shape. The figure 5 shows a schema of the device. Figure 5: basic pneumatic flexible element Assuming an adiabatic compression yield: γγ = xxii VpVp (5.2) px and Vx are the absolute pressure and volume after the piston moved by x. The pressure and the volume can be expressed as below: AxVV, A Fp ix −== (5.2,3) In assuming that the piston surface area is constant, the stiffness of the pneumatic flexible element is finally: 1 i i 2 i 0 x V A1 1 V Ap dx dFk +γ                     − γ == (5.4) If Ax<>== The stiffness is then approximately proportional to the weight of the supported object; the pneumatic flexible element has a natural “constant natural frequency” characteristic. The equation (5.4) will be used as a model of the air volume stiffness of the pneumatic mount. 3. Single action pneumatic flexible element with damping chamber Another more complicated passive pneumatic vibration isolator is a self-damped pneumatic spring (which itself has no significant damping) in which the main cylinder is connected with a damping chamber through a capillary. This device is described in [7] by I. Rivin. It is shown in figure 6. x F m p,V 13 Figure 6: Single action pneumatic flexible element with damping chamber The flow through a capillary, for a passage whose length is much greater than its cross sectional dimension, e.g. the diameter, is: ( )2 2 2 1 0 c 1t pp RT2 gCW −= (5.6) which can be linearized as: ( )21 0 ci 1t pp RT gCpW −= (5.7) since for small oscillations and laminar flow: i21 p2pp ≈+ The Reynolds number is able to give an idea about the fluid behavior. The flow can be assumed linear if the Reynolds number Re< 2000 (which corresponds to a low fluid flow velocity and/or a high viscosity) or, in taking a prudent safety margin, Re<500. Since the weight low rate from the damping chamber 2 can be expressed as: dt dp RT gVW 3 0 3 1t γ −= (5.8) Wt1 can be recalculated in putting the expression of p3 from the equation (5.7) into (5.8), which gives:       + γ −= dt dp dt dW Cgp RT RT gVW 11t ci 0 0 3 1t (5.9) The weight flow rate into the chamber 1 is:       + γ −= dt dVp dt dpV RT gW 1t 1 11t 0 1t (5.10) Since one can approximates: AxVV,Ppp i1i1 −=∆+= and that one can assume: Ax<10, can be applied. The impedance can be then reduced to the more simple expression: ( ) 2i r j1 r 2jZ π       + µβ +ωρ= l (7.6) Since in our case, for 30 Hz, 6.3kr =α= , the equation (7.6) can be used to build the model parameters. From [12], the inner impedance of a small orifice whose length and diameter are comparable in size can be written as: ( ) ( ) A v 5.0 )j2(rkJ )j2(rkJ )j2(rk 21 1j r Z 21 20 21 21 21 2 2i ρ + − − − − ωρ π = l (7.7) This is actually the same expression than the previous one but including a non linear term A v 5.0 ρ taking into account that the impedance is increasing with the air particle velocity, v, in the hole. In fact, in [12], the real part of the impedance is plotted for different speed flow in the orifice and it is pointed out that the nonlinearity that can be seen is a velocity effect rather than amplitude or acceleration effect. 22 3. Radiation impedance The radiation mechanical impedance of a plane piston on infinite baffle, ZMr, is defined by Beranek in Ref [5]. The acoustical impedance ZAr is then: 22 MrAr )r(ZZ π= ;       +− π ρ = rk )rk2(Hj rk )rk2(J1 r cZ 3 31 3 31 2Ar (7.8) In the formula above, k3=ω/c is the wave number. J1 is the Bessel function of first order and H1 is the Struve function of order one. Since there is no readily available Struve function in MATLAB, one has to approximate the Struve function with the firsts terms of the power series expansion:       −+− π = ... 7531 z 531 z 31 z2)z(H 222 6 22 4 2 2 1 Another approximation could also be, from 14: 201 z )zcos(13612 z )zsin(516)z(J2)z(H −       π −+      − π +− π ≈ But the difference occurs at high frequency and the frequency range we are working on is from 0 to 30 Hz; then the final result is not changed and either of these expressions could be used. In the figure 10 below is drawn the real and imaginary part of the mechanical impedance of a piston in infinite baffle in logarithmic scale as in Ref [5]. Figure 10:Real and imaginary part of the normalized mechanical impedance of the air load on one side of a plane piston mounted in an infinite flat baffle 23 VI. Electro- mechanical-acoustical analogy It could be useful to make an electro-mechanical-acoustical analogy, which means applying the electrical-circuit theory to the solution of mechanical and acoustical problems. In [5] an electrical analogy for mechanical and acoustic mass, compliance and resistance are largely discussed. In fact, they are the basic element that should be combined in order to get more complicated devices. That is why, with the help of [5] the impedance-type analogy for acoustical mass, acoustical compliance and acoustical resistance will be shown here in the table 4 below. To transform a physical low into acoustical terms, one uses the formula: S )t(f)t(p = where p is the pressure, f the force and S the section area. 24 Eq ua tio ns A co us tic al el em en t U ni t Im pe da nc e- ty pe an al og y an d sy m bo l Ph ys ic al la w c on ve rte d in to a co us tic al te rm s: St ea dy st at e ( f 2π = ω ) A co us tic al e le m en t M as s K g/ m In du ct an ce (L ) dt )t( dU M )t(p A = U M j p A ω = Tu be fi lle d w ith g as o f l en gt h l a nd c ro ss - se ct io na l a re a S. A co us tic al co m pl ia nc e m /N C ap ac ita nc e (C E) ∫ = dt)t( U C1 )t(p A A C j U p ω = En cl os ed v ol um e of a ir V w ith o pe ni ng fo r e nt ra nc e of p re ss ur e va ria tio ns . A co us tic al re si st an ce N /s m R es is ta nc e (R E) )t( U R )t(p A = U R p A = Fi ne -m es h sc re en A ny im pe da nc e Ta bl e 4: A co us tic al -E le ct ri ca l a na lo gi es U C A p U M A p Z A U p l V ol um e V o f ai r Up Z = 25 VII. The measurements Here is explained how the measurements have been done and the interpretation of them in order to build the model. 1. General set-up The dynamics measurements are done on a servo-hydraulic machinery (MTS-1000 Hz). This machinery is behaving rigid up to high frequencies (1000 Hz at least) in order to avoid bending waves in the set-up. The machinery can do measurements up to 1000 Hz. Figure 11: Measurement set-up: the servo-hydraulic machinery The displacement is given and measured at the upper side of the piston of the mount. The resulting force at the steal plate is measured. With those data, the stiffness can be calculated in two ways: • Calculate the peak-to-peak value of the force and displacement signal, which gives the absolute value of the dynamic stiffness. The phase is calculated out of the actual hysteriesis curve area. • The second method is to use the fast Fourier transform. The amplitude of the fundamental wave is calculated. The phase shift is the difference of the load phase and the displacement phase. To validate the model, different measurements are done. Three preloads are chosen: -850 N, -1000 N and -1200 N (it covers different type of vehicle in which the damper could be mount in) and two excitation amplitude: 1 mm and 0.1 mm (it covers different driving conditions and different roads types). The chosen frequency range is from 0 to 30 Hz. 26 The dynamic stiffness measurements are first done on the mount described on chapter IV. Then, the dynamic stiffness measurements are also done for a closed orifice and for an orifice of 20 mm of diameter in order to test the total mount highest and lower stiffness limits. In all the following plots, the frequency is in Hertz, the absolute value of the stiffness in N/m, and the angle in degrees. 2. Static measurement The static measurements are done in a spindle machinery. The speed given is then very well controlled. Since this machinery is not completely stiff, a sensor is measuring the force due to this lack of stiffness and the results are then more accurate by updating with the sensor data. -14 -12 -10 -8 -6 -4 -2 0 2 -4000 -3000 -2000 -1000 0 1000 2000 3000 4000 Displacement (mm) Z axis Fo rc e (N ) Z a xi s Figure 12: Force-displacement curve 2 mm 1000 N 27 3. Dynamic measurements and results 0 5 10 15 20 25 30 1 1.5 2 2.5 3 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k -850 N -1000 N -1200 N Figure 13: Measured dynamic stiffness and phase angle for 0.1 mm amplitude excitation 0 5 10 15 20 25 30 1 1.5 2 2.5 3 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k -850 N -1000 N -1200 N Figure 14: Measured dynamic stiffness and phase angle for 1 mm amplitude excitation Out of those measurements (figure 13 and 14), one can see that the amplitude of excitation plays a big role. In plotting the curves of the absolute value and the phase angle of the complex stiffness for different preload on the same graph, one can also see a preload influence. The preload has an influence on the working chamber volume. In fact, if the preload increases, the chamber volume decreases, which result in a lower compressibility of the air enclosed volume and to higher stiffness. But the shift of the absolute value to lower values when the preload increases is also due to the rubber dependence on preload, which is not modelled. 28 4. Pneumatic engine mount with closed orifice Before closing the orifice, a deflection is imposed to the mount. A deflection of 5.5 mm corresponds to –850 N preload, a deflection of 6.5 mm corresponds to –1000 N preload, and finally, a deflection of 7.5 mm corresponds to –1200 preload. Then, after closing the orifice, the pressure inside the mount is the atmospheric pressure. Thereafter, the dynamic measurements can be done. 0 5 10 15 20 25 30 1 1.5 2 2.5 3 x 105 frequency ab s k 0 5 10 15 20 25 30 0 1 2 3 4 frequency an gl e k -850 -1000 -1200 Figure 15: Measured dynamic stiffness and phase angle for 1 mm amplitude excitation; closed orifice configuration 0 5 10 15 20 25 30 1 1.5 2 2.5 3 x 105 frequency ab s k 0 5 10 15 20 25 30 0 1 2 3 4 frequency an gl e k -850 -1000 -1200 Figure 16: Measured dynamic stiffness and phase angle for 0.1 mm amplitude excitation; closed orifice configuration Out of those results (figures 15 and 16), one gets the dynamic stiffness contribution of the air chamber and the rubber. This measurement gives the upper limit of the dynamic stiffness of the mount. What can also be remarked, is that the stiffness for –1200 N preload is larger than for the other two preloads. In fact, the air volume effect is here seen and at –1200 N the working chamber volume is 29 much smaller than for the other preloads and the change of pressure over the change of volume is much higher. That is why the stiffness is larger for –1200 N preload. 5. The pneumatic engine mount with opened orifice The orifice is opened in order to be 20 mm diameter. Then the dynamic stiffness measurements are done again. 0 5 10 15 20 25 30 1 1.5 2 2.5 3 x 105 frequency ab s k 0 5 10 15 20 25 30 0 1 2 3 4 frequency an gl e k -850 -1000 -1200 Figure 17: Measured dynamic stiffness and phase angle for 1 mm amplitude excitation; opened orifice configuration 0 5 10 15 20 25 30 1 1.5 2 2.5 3 x 105 frequency ab s k 0 5 10 15 20 25 30 0 1 2 3 4 frequency an gl e k -850 -1000 -1200 Figure 18: Measured dynamic stiffness and phase angle for 0.1 mm amplitude excitation; opened orifice configuration Out of those results (curves 17 and 18), one gets the rubber dynamic properties in the proper geometry: the elastic stiffness and phase angle. In fact, the air is not playing any role in this configuration since the air volume is seen as infinite since the orifice is big: the air stiffness is much lower than the rubber stiffness. In this case, the total mount stiffness for –850 N preload is the highest because the slope of the force deflection curve, which gives the static stiffness of the mount, is decreasing then increasing slightly (see figure 10). 30 VIII. Description of the model In this part will be derived the equation of the total dynamic complex stiffness of the pneumatic engine mount, which will be used to implement the model of the pneumatic engine mount with MATLAB. 1. Unknowns and assumptions Figure 19: mount schema Ap Cair dr F l Pout Pv q r S V xe Zi Zr Piston area, [m2] Air compliance, [m5/N] Rubber viscous damping, [N/(m.s)] Excitation force, [N] Thickness of the plate (length of the orifice), [m] Pressure at the output of the hole, [Pa] Chamber pressure, [Pa] Air flow through the hole, [m3/s] Radius of the orifice, [m] Steel plate area, [m2] Chamber volume, [m3] Displacement amplitude, [m] Inner impedance of the hole, [Ns/m5] Radiation impedance of the hole, [Ns/m5] δ Rubber loss angle, [°] κair+hole Orifice and air stiffness contribution, [N/m] κr Rubber linear stiffness, [N/m] κrtot Complex rubber stiffness, [N/m] κtot Total mount stiffness, [N/m] Ap Pout, q pv x F 31 The different parts of the model to take into account are: • The rubber, which is modelled with Voigt model. • The hole, which is modelled has having an inner impedance and an output impedance. Those impedances are both defined in chapter VI. • The air volume, which is modelled has having a stiffness of a pneumatic spring as in chapter IV. The unknowns of the model are: Pv, Pout, q. It is assumed that: • The density of the gas in the chamber is constant in space and time • The equations are transformed into Laplace domain to get the transfer complex stiffness 2. The air compliance expression The air compliance κair is computed as a closed volume compliance, which is the same as in the paragraph 2 of chapter IV. )1( e 0 p0 2 pin air x V A 1 1 V Ap +γ                     − γ =κ (10.1) 3. The rubber model used The rubber is modelled with Voigt model. For this model the rubber is taken has having a linear stiffness κr and a viscous damping dr. Figure 20: Voigt model of rubber element κr dr x F 32 The differential equation is then: fxxd rr =κ+& , and in frequency domain: FX)dj( rr =κ+ω The loss angle of the rubber is defined as: r rd)tan( κ ω =δ and finally: ( ))tan(j1rrtot δ⋅+κ=κ (10.2) 4. The basic equations used for the fluid The equations concerning the air are: The force equilibrium of the air: )t(Sp)t(F v= The continuity equation in term of air compliance: air v C V)t(p ∆ = , air 2 p air A C κ = The volume conservation: tiep AxxAV ∆−=∆ The equation concerning the hole: The driving flow through the hole: )t(qZ)t(p)t(p tioutv =− The definition of the radiation impedance of the hole: )t(qZ)t(p tsout = (10.3) (10.4) (10.5) (10.6) (10.7) κair is the air compliance as defined in equation (10.1). Zi and Zs are defined in chapter VII. Then:       +ω +=∆ )ZZ(j pxAV si v ep airsi v ep air v C 1 )ZZ(j pxA C Vp       +ω −= ∆ =       +ω + = airsi airep v C)ZZ(j 11 CxA p Finally:       +ω + ==κ + airsi airp e holeair C)ZZ(j 11 CSA x F (10.8) Since the air and the rubber spring are playing in parallel, the total complex stiffness is: ( ))tan(j1 C)ZZ(j 11 CSA r airsi airp rtotholeairtot δ⋅+κ+       +ω + =κ+κ=κ + (10.9) 33 Then, the term κair+orifice can be written as: airpsip airsi 2 p CSA)ZZ(jSA C)ZZ(j)SA( ++ω +ω . If Zs is neglected and Zi expressed as in equation (7.6), on gets the equation (7.8) ( ) ( ) airp p p air p 2 p airpsip airsi 2 p holeair CSA A j1 r 2jjSA C A j1 r 2jj)SA( CSA)ZZ(jSA C)ZZ(j)SA( +      + µβ +ωρω       + µβ +ωρω = ++ω +ω =κ + l l (7.8) Finally: ( ) ( ) airp 2 airp 2 holeair CSAj1j r 2)j(S CSAj1j r 2)j(S +      +ω µβ +ρω ⋅      +ω µβ +ρω =κ + l l The air volume is modelled as a spring with a stiffness κair-model= airp CSA and the orifice is modelled as a mass lSmorifice ρ= in parallel with a dashpot ( )j1 r 2dorifice + µβ = . The air volume and orifice are acting in serie. The analogous mechanical system is in figure 21. Figure 21: pneumatic mount model analogous mechanical system The mass morifice corresponds to the mass of air in the orifice dorifice corresponds to the losses in the hole due to shear forces. 5. The limits of the model The upper limit is given by the rubber stiffness together with the air stiffness for closed volume and the lower limit is given by the rubber stiffness only, as pointed out with the measurements. 6. Numerical implementation in MATLAB In the inner impedance non linear term A v 5.0 ρ (equation (7.7)), the particle velocity v has to be first calculated as exv ω= as a rough approximation (this formula is true for compressible fluid). x F κair κr dr mair dair x F κair-model κr dr morifice dorifice 34 This expression will be updated as: )ZZ(C1 ASxjv siair te ++ ω = . The last formula comes from equation (10.6). Figure 22: Program functionment chart The volume and the pneumatic diameter are calculated with finite element method for the different preload. The preload influence is taken into account through the initial volume and the pneumatic diameter. 7. Parameter study A parameter study in which the absolute value and the phase angle of the stiffness have been plotted in the same graph for a variation of ± 10 % of each parameter has been done (see appendix B). The parameters are: the length of the hole, the diameter of the hole, the pneumatic diameter, the chamber volume, the steal plate diameter, the rubber stiffness and phase angle. teold Axq ω= iZ totκ If 000001.0qq oldnew >− newold qq = NO YES totκ calculation and graph plot qnew 35 The conclusions of it are that: • The diameter of the orifice gives the resonance frequency and can be adapted to tune the mount. • The air chamber volume should be decreased in order to increase the damping. It also determines the behavior at the highest frequencies. • The pneumatic diameter and the diameter of the steel plate below should be increased in order to increase the damping. As the volume, it also determines the dynamic stiffness at the highest frequencies. • The rubber elastic stiffness should be as low as possible to get the highest damping. This confirms the theory used and that the model is working in a proper way. 36 IX. Model/measurements comparison 1. The rubber measurement and model comparison From the Voigt model, the rubber absolute stiffness and phase angle are taken as constant values, which could be chosen as the average of the measured absolute stiffness and phase angle with opened orifice configuration. In that case, the curves are good only after about 10 Hz. In fact, it is hard to model the dynamic properties of the rubber at low frequencies. But those values can also be taken from the measurements with opened hole, and then as frequency dependent. In this case, the curves are fitting with the measured one in the entire frequency range. In the figure 23 are the comparison between the calculated curves with the rubber properties taken as constant and the calculated curves with measured rubber properties. 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 x 105 frequency ab s k kr constant kr meas 0 5 10 15 20 25 30 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 x 105 frequency ab s k kr constant kr meas -850 N preload, 1 mm amplitude -850 N preload, 0.1 mm amplitude Figure 23: Comparison between different rubber properties calculation 2. Measurements and model fit In looking at the parameter study, the relevant parameters that could be tuned are: • The pneumatic diameter • The volume • The rubber elastic stiffness • The rubber phase angle. In fact, the diameter of the orifice is only shifting the resonance frequency of the system, which the model gives correctly and there is no reason that the steal plate diameter should be wrong. Anyway, since the influence of this parameter is the same as the pneumatic diameter, if there is any uncertainty, it would be contained in the pneumatic diameter factor. In the appendix C are plotted the stiffness and phase angle of the mount: in the same figure, the calculated (calc) and measured (meas) curves can be seen. Regarding the curves comparison without factor, it can be seen that the stiffness for 1 Hz are not matching. This should be the case because at low frequencies, the air movement is so low that the behavior of the mount is similar to the static case and only the elastic stiffness of the rubber is seen. But this could by explained by the fact that for the calculates curves, the rubber stiffness is taken from the opened orifice measurements, which gives a slightly different rubber elastic stiffness than 37 in the normal dynamic measurements: the rubber properties are changing with the time and are sensible to fatigue. Then, if the calculated curves are shifted to the same start stiffness value as the measured one, the calculated absolute value of the stiffness is getting higher than the measured one for higher frequencies. The phenomenon is the same for the phase angle. It can be concluded that the calculated air stiffness is too high, which is confirmed by the fact that the volume has to be multiplied by a factor almost two in order to get a good fit between the measured and calculated curves (the total mount stiffness is depending on the volume only via the air volume stiffness. There are only some clues to explain this: • The volume is really small and some unknown phenomena could happen. • The air stiffness is calculated for a squared box geometry: the geometry of the rubber part of the mount is not taken into account and could play a role • There could be some measurements uncertainty. 38 X. Proposals how to increase the damping 1. Several holes, orifice area increase With the model, it could be shown that having several small holes (smaller than the one taken so far), increases slightly the damping. In the same graph are plotted the absolute value and the phase angle of the complex stiffness for different areas: • The first area is the normal orifice area corresponding to a circular orifice of 2 mm diameter • The second area correspond to two circular orifices of 2 mm diameter • The third area correspond to two orifices of 1 mm diameter • The forth area correspond to four orifices of 2/1 mm diameter The first, third and fourth case corresponds to a same total area. 0 5 10 15 20 25 30 1.6 1.7 1.8 1.9 2 2.1 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k area normal 2 holes 2 holes diam/sqrt(2) 4 holes diam/2 Figure 22: Computed dynamic stiffness and phase angle for 0.1 mm amplitude excitation; different orifice configurations 39 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k area normal 2 holes 2 holes diam/sqrt(2) 4 holes diam/2 Figure 23: Computed dynamic stiffness and phase angle for 1 mm amplitude excitation; different orifice configurations Then it can be seen that increasing the total area is shifting the resonance frequency as pointed in the parameter study, but for a same area distributed on several holes, the damping is almost not increasing and since it would anyway be too complicated to make it industrially, no measurements has been done to verify those results. 2. Acoustic absorbent inside the volume L. Beranek in Ref [5] is discussing the influence of absorptive lining in loudspeaker closed box. A conclusion is that the absorber makes the speed of sound decreasing in the box because the compression becomes isothermal. Then, the ratio of specific heats decreases from 1.4 to 1, which means that the air compliance is increasing. Since it is difficult to get the absorber acoustical properties, it hasn’t been tried to include them in the model. 40 XI. Conclusion and discussion After having understood how an hydraulic mount is working and explored some simple pneumatic spring, the model of the pneumatic mount has been implemented in MATLAB and finally, the calculated curves are compared to the measured one. The mount has physically been divided in three parts in order to model it more easily and the conclusion of the model-experiment study shows that: • The rubber part is well modelled with a spring in parallel with a dashpot, • The orifice impedances seems to be quite correctly modelled since the global model predicts really well the resonance frequency; in fact, the channel geometry and behavior determine the frequency of maximum phase angle and the damping properties. • The air compliance is not modelled accurately enough since the calculated phase angle is always too high compared to the measured one. It can be also remarked that the simplified orifice impedance is working well for the chosen frequency range and for this application. Then, the next step to do in order to improve the model would be to find a better air stiffness approximation: the total mount stiffness at the highest frequencies would be better modelled. For that, the air stiffness model should include the mount geometry. Finally, a non-linear model for large amplitudes with small air cavity volume could be useful to get a modelisation in time domain. 41 REFERENCES 1. J. E. COLGATE, C.-T. CHANG, Y.-C. CHIOU, W. K. LIU AND L. M. KEER 1995. Modelling of a hydraulic engine mount focusing on response to sinusoidal and composite excitations, Journal of Sound and Vibration, 184:503-528. 2. R. SINGH, G. KIM AND P. V. RAVINDRA 1992. Linear analysis of automotive hydro- mechanical mount with emphasis on decoupler characteristics, Journal of Sound and Vibration, 158:219-243. 3. SHI-JIAN ZHU, XUE-TAO WENG, GANG CHEN 2003. Modelling of the stiffness of elastic body, Journal of Sound and Vibration, 262:1-9. 4. CLARENCE W. DE SILVA 1999. Vibration: fundamental and practice, CRC press. 5. LEO L. BERANEK 1954. Acoustics, Mac Graw Hill, New-York. 6. J.A. FOX 1974. An introduction to engineering fluid mechanics. 7. EUGENE I. RIVIN 2003. Passive Vibration Isolation, ASME press. 8. M. SORLI, L. GASTALDI, E. CODINA, S. DE LAS HERAS 1999. Dynamic analysis of pneumatic actuators, Simulation Practice and Theory 7, 589-602. 9. W.M.BELTMAN 1998. Viscothermal wave propagation including acousto-elastic interaction, PhD thesis, University of Twente, The Netherland. 10. M. HOFMANN 2002. Antivibration systems. 11. A. GEISBERGER, A. KHAJEPOUR AND F. GOLNARAGHI 2002. Non-linear modelling of hydraulic mount: theory and experiment, Journal of Sound and Vibration, 249:371-397. 12. L. J. SIVIAN 1935, Acoustic impedance of small orifices, Journal of the acoustical society of America, volume 7:94-101. 13. B. I. BACHRACH and E. RIVIN 1982, Analysis of a damped pneumatic spring, Journal of Sound and Vibration, 86(2):191-197. 14. RONALD M. AARTS and AUGUSTUS J. E. M. JANSSEN 2003, Approximation of the struve function H occurring in impedance calculations, Journal of the acoustical society of America, 113 (5): 2635-2637 15. IRVING B. CRANDALL 1927. Theory of vibrating systems and sounds, D. Van Nostrand Company, New-York. 16. Engineering with Rubber; edited by Alan N. GENT with contributions by R.P. CAMPION and others, University of Akron, USA 42 XII. APPENDIX A1: Parameter study for –850 N preload and 0.1 mm amplitude Volume variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k V0 normal V0-10% V0+10% 43 Length of the orifice variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k length normal lt-10% lt+10% Diameter of the orifice variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k diameter normal dt-10 dt+10% 44 Angle of rubber damping variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k angle normal angle-10% angle+10% Rubber elastic stiffness variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k kr normal kr-10% kr+10% 45 Diameter of the plate below variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k plate normal dp-10% dp+10% Pneumatic diameter variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k pd normal pd-10% pd+10% 46 XIII. APPENDIX A2: Parameter study for –850 N preload and 1 mm amplitude Volume variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k V0 normal V0-10% V0+10% 47 Length of the orifice variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k length normal lt-10% lt+10% Diameter of the orifice variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k diameter normal dt-10% dt+10% 48 Angle of rubber damping variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k angle normal angle-10% angle+10% Rubber elastic stiffness variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k kr normal kr-10% kr+10% 49 Diameter of the plate below variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k plate normal dp-10% dp+10% Pneumatic diameter variation 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k 0 5 10 15 20 25 30 0 5 10 15 20 frequency an gl e k pd normal pd-10% pd+10% 50 XIV. APPENDIX B: Calculated and measured curves comparison -850 N preload, 0.1 mm amplitude With coefficient Without coefficient Absolute value 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 10 5 frequency ab s k calc meas 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 10 5 frequency ab s k calc meas Phase angle 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 20 frequency an gl e k calc meas 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 20 frequency an gl e k calc meas 51 -850 N preload, 1 mm amplitude With coefficient Without coefficient Absolute value 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas Phase angle 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 20 frequency an gl e k calc meas 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 20 frequency an gl e k calc meas 52 -1000 N preload, 0.1 mm amplitude With coefficient Without coefficient Absolute value 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas Phase angle 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k calc meas 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k calc meas 53 -1000 N preload, 1 mm amplitude With coefficient Without coefficient Absolute value 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas Phase angle 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 20 frequency an gl e k calc meas 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 20 frequency an gl e k calc meas 54 -1200 N preload, 0.1 mm amplitude With coefficient Without coefficient Absolute value 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 x 105 frequency ab s k calc meas Phase angle 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k calc meas 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k calc meas 55 -1200 N preload, 1mm amplitude With coefficient Without coefficient Absolute value 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k calc meas 0 5 10 15 20 25 30 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 105 frequency ab s k calc meas Phase angle 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k calc meas 0 5 10 15 20 25 30 0 5 10 15 20 25 frequency an gl e k calc meas 56 XV. APPENDIX C: MATLAB code clear %GEOMETRY pd=0.08605; %pneumatic diameter m; -850 N % pd=0.08622; %pneumatic diameter m; -1000 N % pd=0.08619*0.93; %pneumatic diameter m; -1200 N Ap=pi*(pd/2)^2; %piston area m^2 dt=0.002; %diameter of the hole m r=dt/2; %radius of the hole m At=pi*r^2; %cross section area of the hole m^2 dp=0.1017; %diameter of the plate below m S=pi*(dp/2)^2 %area of the plate below m^2 lt=0.002; %lenght of the hole m V0=38.4878*10^(-6); %initial volume m^3; -850 N % V0=33.100*10^(-6); %initial volume m^3; -1000 N % V0=25.627*10^(-6)*1.8; %initial volume m^3; -1200 N %CONSTANTS p0=101300 %atmospherique pressure Pa rho=1.2255;%kg/m^3 gamma=1; %ratio of specific heats kinvisc=17.98*10^(-6);%fluid dynamic viscosity (Pa.s) c=343;%speed of sound in air m/s %PARAMETERS xe=0.001;%amplitude of the displacement % xe=0.0001;%amplitude of the displacement %AIR COMPLIANCE CALCULUS aircomp=V0./(((1-Ap*xe/V0).^(gamma+1)).*(gamma*p0)); 57 %VECTORS omega=linspace(1,30,30)*2*pi; s=j*omega; ki=(omega.*rho./(2*kinvisc)).^(0.5); k=omega./c; wave=k.*r; omegameas=linspace(1,30,30)*2*pi; bessel_1=2./pi.*(((2.*wave).^3./3)-(2.*wave).^5./(3.^2.*5)+(2.*wave).^7./(3.^2.*5.^2.*7)- (2*wave).^9/(3.^2.*5.^2.*7^2*9)); %H1 serie bessel_2=2./pi-besselj(0,2.*wave)+(16./pi-5).*(sin(2.*wave)./(2.*wave))+(12-36./pi).*(1- cos(2.*wave))./((2.*wave).^2); Zsunnorm=(rho.*c.*(pi.*r.^2)).*((1-besselj(1,2.*wave)./(wave))+j.*bessel_1./wave); % plane piston in infinite baffle Zs=Zsunnorm./((pi*r^2)^2); Ziwf=(lt./(pi.*r.^2))*s.*rho.*(1./(1-(2./(ki.*r.*(-2*j).^(0.5))).*besselj(1,ki.*r*(- 2*j).^(0.5))./besselj(0,ki.*r.*(-2*j).^(0.5)))); %internal impedance of a small operture for m=30:-1:1 err=1; qiold(m)=xe*Ap*omega(m); load measurefermeouvert kr(m)=Kmeasferme2(m)*10^3*1.13*(1+j.*tan(pi*anglemeasferme2(m)*1.1/180)); while err>0.000001 Zi(m)=Ziwf(m)+0.5*rho*abs(qiold(m))/At^2; qi(m)=(xe*s(m)*Ap)/(1+((aircomp)*s(m)*(Zi(m)+Zs(m)))); K(m)=(S*Ap/(aircomp))/(1+1/(s(m)*(Zi(m)+Zs(m))*(aircomp)))+kr(m); err=abs(qi(m)-qiold(m)); qiold(m)=qi(m); end 58 end load measured_values_850_mine figure(1) plot(omega/(2*pi),abs(K),... omegameas/(2*pi),Kmeas2*10^3) xlabel('frequency') ylabel('abs k') axis([0 30 1.3e5 3e5]) figure(2) plot(omega/(2*pi),angle(K)*180/pi,... omegameas/(2*pi),anglemeas2) xlabel('frequency') ylabel('angle k') legend('calc','meas') axis([0 30 0 20])