Transcript
Acoustical properties for diesel particulate filters
LINUS FALK LISSEL JESPER KRISTOFFERSSON
SA105X Examensarbete på kandidatnivå i ljud, vibrationer och signaler
Stockholm 2011
Acoustical properties for diesel particulate filters
Marcus Wallenberg Laboratoriet
Linus Falk Lissel Jesper Kristoffersson
Datum: 2011‐05‐19 Ansvarig handledare: Mats Åbom
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Abstract Diesel Particulate Filter (DPF) is still not mandatory for trucks and busses but will in a near future probably be so. This is one of the reasons why many manufacturers are interested in how a DPF affects the performance of their vehicles. This report investigates the transmission loss and pressure drop caused by the DPF. The pressure drop is of interest because it affects the performance and fuel consumption. Due to the fact that a DPF has sound‐absorbing properties, it can be incorporated in the muffler system. Four DPF: s were tested of which one was clean and the three others had different degrees of soot layer thicknesses. The filters were placed in the flow test rig at the Marcus Wallenberg Laboratory (MWL) at KTH and the transmission loss and pressure drop were measured. The transmission loss was measured without flow for the filters with soot. This was necessary to minimize the risk of destroying the sensitive microphone membranes. For the clean filter the transmission loss was measured with two different flow speeds (0.05 Mach and 0.1 Mach). The pressure drop was measured on all filters with flow speeds up to 62 m/s. The pressure drop was then plotted against the flow speed for each filter. The transmission loss was measured with the two‐microphone method (TMM). The employed frequency range was 50‐ 1000 Hz and a stepped sinus with steps of 10 Hz was used. The two‐ports for the filters together with the expansion chamber were calculated with SIDLAB. The results from the calculated transmission loss and pressure drop are presented as figures and as tables. The conclusions from the analyses are:
the transmission loss, for the complete system, increases as the soot layer increases, the transmission loss are hardly affected by the flow in the ducts, hence doing the measurements without flow gives accurate results for the investigated frequency range and test object, the pressure drop increases as the soot layer increases. the SIDLAB‐simulation of the transmission loss for the system, is only valid up until 661.5 Hz, this is because that is where the cut‐on frequency for the system is located.
The report also includes a part where the measurement methods and theory behind them are described as well as a study of the available literature on the subjects. The studied articles treat models and experiments on DPF: s regarding transmission loss, acoustical two‐port and pressure drop. Further studies must be done to calculate the transmission loss and the two‐port for each DPF separately, i.e. without the expansion chamber.
2
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Sammanfattning Dieselpartikelfilter (DPF) är ännu inte ett krav för dieseldrivna lastbilar och bussar men kommer troligen bli det inom en snar framtid, detta är en av anledningarna till varför fordonsproducenter intresserar sig för hur dieselpartikelfilter inverkar på ett fordon. Det som presenteras i denna rapport är hur olika sotinlagringsgrad påverkar transmissionsisolering och tryckfall för ett DPF. Tryckfallet är intressant eftersom det påverkar motorns prestanda och bränsleförbrukning. Eftersom ett DPF har ljuddämpande egenskaper kan detta inkorporeras i ljuddämparsystemet för att på så sätt uppnå en bättre total ljuddämpning. Fyra DPF testades varav ett var rent och de tre övriga hade olika grad av sotinlagring. Dessa placerades i flödesriggen i Marcus Wallenberg Laboratoriet (MWL) på KTH, där transmissionsisoleringen och tryckfallet mättes upp. För filterna med sotinlagring mättes transmissionsisoleringen endast utan flöde. Denna begränsning var nödvändig för att minimera risken för att skada de känsliga mikrofonerna. Eftersom flödet inte nämnvärt påverkar transmissionsisoleringen var detta ingen begränsning för resultatet. För det rena filtret mättes transmissionsisoleringen med flöde (0.05M och 0.1M). Tryckfallet mättes för samtliga filter för strömningshastigheter upp till 62 m/s. Tryckfallet plottades därefter mot strömningshastigheten för respektive filter. Transmissionsisoleringen mättes upp med två‐mikrofonsmetoden (TMM). Det använda frekvensomfånget var 50‐1000 Hz och stegad sinus användes med steg om 10 Hz. De akustiska två‐portarna för de olika filterna med tillhörande expansionskammare har tagits fram med hjälp av SIDLAB. Transmissionsisoleringen och tryckfallet presenteras i grafer, och tryckfallet även i en tabell. Slutsatserna efter analyserna är att:
transmissionsisoleringen, för hela systemet ökar, då sotinlagringen ökar, transmissionsisoleringen påverkas endast marginellt av flöde i rören, varvid mätningar utan flöde ger tillfredställande resultat för det undersökta frekvensomfånget och testobjekten, tryckfallet ökar, då sotinlagringen ökar. simuleringen i SIDLAB för transmissionsisoleringen för hela systemet är endast giltig för frekvenser upp till cut‐on frekvensen för systemet, 661.5 Hz
Rapporten innehåller även avsnitt där de använda mätmetoderna och teorin bakom dessa beskrivs samt en litteraturstudie av aktuella forskningsartiklar inom området. De studerade artiklarna behandlar modeller och experiment för DPF: er angående transmissionsisolering, akustisk två‐port och tryckfall.
3
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Ytterligare studier måste göras för att beräkna transmissionsisoleringen och två‐porten för endast filtret, det vill säga utan expansionskammare.
Acknowledgements This bachelor project was carried out during the spring term 2011 as a part of the Bachelor’s programme in Vehicle Engineering at KTH. The task was executed for the Acoustic Department, Scania CV AB in Södertälje in collaboration with the Marcus Wallenberg Laboratory (MWL) at KTH in Stockholm. The examiner was Hans Bodén, Dr. Tech., Department of Aeronautical and Vehicle Engineering, KTH Stockholm. Our tutor at KTH was Mats Åbom, D.Sc. (Tech.) and the supervisor at Scania was Tony Karlsson. We are particularly grateful to Mats Åbom and Dr. Sabry Allam without whose patience, support and thoughts this project would not have been finished within the given time limit. We also would like to thank Danilo Prelevic, Research Technician at MWL, who has been very helpful before, during and after the practical work in the laboratory as well as Sara Elsaadany for the first hand SIDLAB knowledge. Finally we would like to thank all the staff at MWL for their contribution and support in our pursuit for knowledge.
4
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Table of contents Abstract ....................................................................................................................................... 2 Sammanfattning............................................................................................................................ 3 Acknowledgements....................................................................................................................... 4 Table of contents........................................................................................................................... 5 1. Introduction............................................................................................................................... 7 2. Collection of information .......................................................................................................... 8
2.1. Acoustical models........................................................................................................... 8
2.1.1. Measurement technique ..................................................................................... 8
2.1.2. The models........................................................................................................... 9
2.1.3. The lumped impedance model .......................................................................... 10
2.1.4. The 1‐D model, the modified 1‐D model and the new model........................... 10
2.1.5. The 3‐D FEM model ........................................................................................... 13
2.2. Pressure drop models................................................................................................... 14
2.2.1. 1‐D Pressure Drop Model .................................................................................. 15
3. Theory and measurement techniques .................................................................................... 18
3.1. Two‐port ....................................................................................................................... 18
3.2. Microphone calibration ................................................................................................ 19
3.3. Pressure drop ............................................................................................................... 20
3.4. Limitations .................................................................................................................... 20
4. Measuring................................................................................................................................ 22 5. Results .................................................................................................................................... 24
Transmission loss ......................................................................................................... 24
Pressure drop............................................................................................................... 26
6. Discussion................................................................................................................................ 28 7. Conclusion ............................................................................................................................... 30 8. References............................................................................................................................... 31
5
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Appendix A – Filter data Appendix B – Dimensions of the test rig Appendix C – Data sheet Bruel & Kjaer, Nexus model 2691‐A‐OS2 Signal Conditioner sn: 2301659 Appendix D – Microphone Preamplifier Type 2670 Appendix E – Data sheet Microphone ¼″ Pressure‐field Microphone — Type 4938, sn: 2323070, 2491541, 2239162, Brüel & Kjaer Appendix F – Calibrator Appendix G – Data sheet SwemaAir 300 Appendix H – Data sheet NAD C370, SR: H0ZC37006007
6
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
1. Introduction 1.1. Background Due to the constantly higher demands for less particle emission from combustion engines, most of the private cars with diesel engines nowadays have a DPF installed in the exhaust system. In the future it will probably be mandatory in trucks as well. The aim of the filter is to reduce the soot particles in the exhaust gas. The filter consists of narrow channels with porous walls through which the exhaust gas is forced, see figure 1, hence the particles will therefore be stuck inside the filter. The increasing soot layer will attenuate the sound propagation in the exhaust system as well as reduce the efficiency of the engine due to the increased pressure drop over the filter. Knowledge of how the soot layer affects the attenuation and the pressure drop is therefore very important in the design of exhaust systems. Within a few years the limit of sound emission for trucks will decrease from the present 80 dB(A) to probably 78 dB(A). Due to the attenuation properties of the DPF, the DPF might be one of the design parameter in the development of more silent trucks.
Figure 1. Photograph of a typical ceramic diesel particulate filter unit with illustration of the flow path through the filter. The flow enters one channel plugged at the downstream end and then passes through the porous walls into the four neighboring channels plugged at the upstream end. At the inlet side of the filter close to 50% of the channels are plugged. A typical width of the square channels is 1–2mm with a wall thickness of a few tenths of a millimeter. Text and figure from [1].
1.2. Purpose The purpose of this project was to investigate how the transmission loss and the pressure drop over a DPF are affected by a soot layer. This was to be done by studies of present research and 7
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
by doing experiments on real filters. Four filters were to be investigated, one clean filter and three with different soot layer thicknesses. For all the filters the two‐port, transmission loss and the pressure drop over the filter were to be determined.
2. Collection of information The collecting of information for the project was initiated with the use of KTHB’s (the library at KTH) search functions, “SAMSÖK”, which include a vast number of databases. To be able to limit the huge amount of hits that searches including search words “diesel particulate filter” together with words as “acoustics”, “two‐port”, “transfer matrix” and “pressure drop” the search was concentrated to the two most important databases Compendex and Inspec. Since starting with the search words “diesel particulate filter” in Compendex produced 1594 hits the search was limited by using the functions in the search engine. The numbers of search hits were limited by using added search word as “pressure drop” and only displaying articles and reports in English. This gave the total amount of 46 hits and these were manually controlled to be able to select the most interesting articles and reports. The same type of search in Inspec gave 33 results. When searching for articles and reports concerning acoustic properties and transfer matrixes, the same data bases were used in the same manner as described above but with the use of the word “acoustics” which in Inspec gave 5 hits and in Compendex, 18 hits.
2.1. Acoustical models To design the after‐treatment devices (ATD: s) in cars, trucks and buses it is very important to use an accurate acoustical model. The ATD consist of a catalytic converter (CC) and a diesel particulate filter (DPF) [3]. The acoustical model for the CC can be found in the literature and will not be treated in this report. Four of the five models that will be presented in this paper have been developed by Sabry Allam and Mats Åbom, both active at the MWL, at the Department of Aeronautical of Vehicle Engineering at KTH in Stockholm. The fifth model is a simplified version of the 3‐D acoustical model of a DPF whose acoustical properties are analyzed with SYSNOISE software and done by Gao Wenzhi, Feng Liming, Niu Wenbo and Wang Hui, all four active at the Tianjin University in Tianjin in China. All the models aim to determine the two‐port matrix and have been validated by experiments. There exist different types of DPF: s but the one with ceramic foam seem to be the best according to [1] and is also the one that is used in the references [1 and 2]. All the filters used in [1 and 2] were filters for private cars and they all had the length of 250 mm and the other dimensions are presented in table 1 in Appendix A. The dimensions used in [5] are unknown. 2.1.1 Measurement technique All the models presented below aim to find the so called acoustical two‐port, i.e. the transfer matrix which describes the relation between the state before and after the DPF. It is a linear system in the frequencies domain as X = TY (1) 8
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
where X and Y are two state vectors ([2x1]‐vectors) and T is the two‐port ([2x2]‐matrix). The most common state variables are the plane wave acoustic pressure, p, and the volume velocity, q. The two‐port matrix can either be determined by calculations or by experiments. The frequencies of interest come from the pulsations from the engine and are therefore low and hence the two‐port theory is appropriate. The measurement technique used in [1‐2] is the so called two microphone method, TMM. In the TMM two or more flush mounted microphones are used to determine the two‐port matrix. The test rig at MWL can be seen in figure 2. When the two‐port matrix is known, the filter transmission loss can be calculated with equations found in the literature.
Figure 2. Layout of the MWL test ring for determination of acoustic two‐port data. Text and figure from [1].
The 3‐D model was validated by experiments in a semi‐anechoic laboratory with a running engine. The engine was running at 1800, 2100 and 2800 rpm with full load and at 1000 rpm with no load. The insertion loss was then calculated by taking the difference between the sound pressure level at the exhaust pipe with and without a DPF. 2.1.2. The models The models presented below mainly have three different approaches to the problem of determine the two‐port. In the first model by Åbom and Allam [1], the lumped impedance model, the DPF is treated as an acoustic resistance and in the 1‐D model and the modified 1‐D model in [1 and 2] the equations of mass and momentum conservations derived from the 1‐D fluid dynamic equation for a DPF (found in reference [6]) are used. For the new model the equation of energy conservation is included. The last model, by Wenzhi and others, is based on the 3‐D wave equation [5]. In all the models, except from the lumped resistance model, the coupling between the porous walls is described via Darcy’s law [1, 2 and 5]. For the models presented by Åbom and Allam in [1 and 2] some basic assumptions were made: only plane (1‐D) acoustic waves were considered, linear acoustics was valid, the Mach number (M) were small (<0.1), temperature and mean flow speed gradients in the axial direction were neglected, the effects of chemical reactions were neglected. 9
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Plane acoustic waves will appear when the frequencies are low and for a combustion engine the pressure pulsation will have a low frequency. None linear effects are small compared to the linear effects, hence the second assumptions is valid. The Mach number in the exhaust system is in the order of 0.3 but the DPF is normally placed in an expansion chamber hence the Mach number will decrease 3‐5 times. The contribution of temperature gradient and mean flow speed gradient is small and can therefore be neglected. The last assumption is valid due to the fact that chemical reactions mainly occur in the regeneration process [1]. For the model derived by Wenzhi and others the following assumptions were made: temperature and mean flow speed gradients were neglected, source with constant pressure, the pressure and velocity are constant in the cross section of the DPF. The first assumption was discussed above and the second is valid due to the fact that the source impedance for the engine is large. The gradients of pressure and velocity are quite small, so they can be approximated as constant [5]. 2.1.3. The lumped impedance model In this model, the DPF is treated as an acoustic resistance and the wave propagation is neglected. The filter transmission loss was calculated and compared to experiments with clean filters at cold conditions (20° C). The model has good agreement with experiments for low frequencies, 200‐300 Hz at 20° C as can be seen in figure 3, and 400‐600 Hz at operating conditions. The filter transmission loss will be independent of the frequencies [1].
Figure 3. Measured and predicted transmission loss for the (RC200/12) filter with M = 0.015 at the filter inlet. o o o, Measured; —, predicted using lumped model; ‐ ‐ ‐, predicted using 1‐D model with measured wall resistance;+++, predicted using 1‐D model (case (ii)). Text and figure from [1].
2.1.4. The 1D model, the modified 1D model and the new model
10
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
To develop a model for the wave propagation in the channels the coupling between them has to be known, i.e. the propagation constant has to be defined. [2] In these models the DPF is divided into five sections, each with its own two‐port matrix, see figure 4. When the two‐port matrix for each section is known, the two‐port matrix for the whole DPF can be calculated via TDPF = TINTITIITIIITOUT. (2) The in‐ and outlet of the filter represent a change in the area and for low frequencies and small Mach number it can be modeled as a lumped impedance model. The sections I and III can be modeled as an end correction added to the in‐ and outlet respectively, this due to their short length (less than 1 cm). Hence it is only the two‐port matrix for the section II that is different between these three models. The main difference between them is the way the viscous and thermal losses are treated [1 and 2].
Figure 4. Cross‐section of a unit cell in a DPF split into five sections each described by an acoustic two‐ port. Note, the filter section (II) is actually an acoustic four‐port but can be reduced to a two‐port due to the hard walls in section I and III. Text and figure from [1].
In the 1‐D model further assumptions were made for the section II: the air is assumed to be an ideal gas the wave propagation is assumed to be isothermal in the channels, and the flow is assumed to be laminar due to the very narrow channels and that the frequencies are low. Solving the equations for mass and momentum conservations gives an expression for the propagation constant that can be used to find the two‐port for this section. Two different procedures are discussed to determine the speed of sound and damping: i. for low frequencies or narrow ducts the speed of sound should be set equal to the isothermal speed of sound due to the fact that viscous effects dominate and the damping is related to the pressure drop for laminar flow, ii. for high frequencies or wide ducts an adiabatic speed of sound should be applied and the damping should depend on the frequency. At 20° C the second case is probably the best estimation of the temperature and damping. But at operating condition, i.e. about 500 ° C, the first case is the best estimation due to a thicker boundary layer. This model has a good agreement with experiments as can be seen in figure 5. The transmission loss was also calculated at different conditions for one DPF to investigate the effect of mean flow and soot layer. The dimensions for the DPF can be seen in table 2 in Appendix A. In the calculations the fluid was assumed to be an ideal gas (air) and the temperature, static pressure and Mach number were assumed to be 700° K, 100 kPa and 0.020 11
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
respectively at the inlet and outlet of the filter. The soot layer was assumed to be 1/10 of the wall thickness and the permeability for the soot layer was assumed to be 1.5 x 10‐14. As can be seen in figure 5, the transmission loss increased significantly with at soot layer but was hardly affected by the mean flow [1].
Figure 5. Predicted transmission loss for the studied filter unit at M = 0.02 and T = 700 °K before the filter inlet. . . . , Predicted for no flow and no soot; , predicted for no flow in the channels and no soot BUT with flow losses at the inlet/outlet; ‐ ‐ ‐ , predicted with flow and no soot; o o o, predicted with flow and soot layer. Text and figure from [1].
But one shortcoming of the 1‐D model is the approximate treatment of the vicious and thermal losses. One way to solve this is to assume a plane thermo‐viscous wave field and use the Kirchhoff solutions for a cylindrical pipe to describe its properties. This is done in the 1‐D model where the viscous friction factor is dropped and the speed of sound and the density are defined in a different way with a Bessel function of the first type [2]. According to reference [3] the modified 1‐D model has three assumptions that limit the model: the effects of the gradient in the mean flow and ambient state are neglected axial gradients are smaller than cross section gradients in the channels the pressure is constant over the cross section in the channels. The first assumption is valid, with exception for the in‐ and outlet, due to the fact that the Mach number is small. The use of the temperature average gives an error of 0.3 dB and can therefore be neglected. The second assumption is not valid in the channels openings where there will be a near field but the effects of this field can be neglected according to the experiments. The pressure can be assumed to be constant if the wall permeability is low. This is valid as long as the wall impedance is much smaller than the impedance for the plane‐wave. This ratio is in the order of 100 for a typical DPF [3]. In the new model the viscous and thermal losses are treated by solving the convective acoustical wave equation with the theory of Zwikker and Kosten. To determine the two‐port matrix for section II, it is divided into segments in which the mean flow and gradients of the 12
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
mean flow are assumed to be constant. The solution will approach the correct solution when the number of segments goes to infinity [2]. As can be seen in figure 6, the modified 1‐D model and the new model show a very good agreement with the experiments.
Figure 6. Measured and predicted transmission loss for the (RC200/12) filter at M = 0.015 before the inlet of the filter and T = 293° K. o o o, Measured; —, predicted using present theory; ……, predicted using the modified 1‐D Model. Text and figure from [2].
After the experiments some conclusions were made: the contribution of mean flow to the transmission loss is low and can therefore be neglected using the average temperature in the channels is a good approximation the soot gives a significant contribution to the transmission loss the 1‐D model gives satisfactorily results for frequencies up to 600‐800 Hz the modified 1‐D model and the new model show a very good agreement with experiments up to 1250 Hz [4]. 2.1.5. The 3D FEM model Due to the complexity of the DPF, the four models in 2.1.3. and 2.1.4.only have a sufficient accuracy for low frequencies up to 1250 Hz. Thanks to the symmetry of the DPF a simplified model can be applied in a FEM‐program where only one quarter of the cross section is treated. The insertion loss was then calculated in SYSNOISE. One advantage with this model over the models developed by Åbom and Allam is that it treats the effects of radial wave propagation. After the experiments some conclusions were made: there is a significant effect of viscosity of the exhaust gas, especially for high frequencies the DPF attenuates the noise significantly, the insertion loss increases for higher frequencies, 13
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
the 3‐D FEM model has a satisfying accuracy for frequencies up to 5000 Hz [5].
2.2. Pressure drop When a diesel particulate filter is inserted in the exhaust system it inevitably creates a pressure drop in some extent. A pressure drop means that some kind of backpressure on the engine is present which forces the engine to work harder and therefore consume more gas which means a lower efficiency. The problem increases as the diesel particle filter gets loaded with soot. To be able to come by the clogging of soot particles the filters are being regenerated. The regeneration process involves oxidation of the accumulated particles and is based either on the use of catalysts to lower the reaction temperature or by electrical heating to increase the temperature of the diesel particulate filter. To determine when to execute an on‐board regeneration process the actual pressure drop could be measured and when reached a predetermined point the regeneration process is initiated. There are different mathematical models that describe the filter pressure drop. The parameters included in the models are the filter geometry parameters (volume, cell density and wall thickness), operating conditions (flow rate and temperature) and the accumulated soot mass. They can be either of analytical character or based on the solution of mass and momentum balances in the filter channel. The latter model requires a computationally demanding solution of a single‐channel model but has greater advantages and is discussed in [7]. Testing pressure drop models is not always the easiest thing to perform, this due to the fact that the diesel engine continuously emits soot. A common approach to solve this is to perform tests with air as the flowing gas at room temperature. In [7], the problem is handled by using two cascading filters after a diesel engine, se figure 7. The first filter is responsible for collecting practically all the soot particles and the pressure drop measurements are then performed on the second filter under different flow rates and temperatures. The accumulated soot mass is then measured by weighing the second filter.
14
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Figure 7. Experimental set-up. Figure and text from [10].
15
Jesper Kristoffersson Linus Falk Lissel
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
2.2.1. 1D Pressure Drop Model To be able to predict the pressure drop in a diesel particulate filter a mathematical model could be used to describe the flow in the filter channels, through the soot and through the filter walls as well as what happens in the entrance and exit see figure 8. Bissett [9] presented a one dimensional model that describes the velocity and pressure variations inside a filter channel. The model relies on the transient solution of the mass, momentum and energy balances in a single channel, with the following major assumptions:
all the filter channels behave in the same way. This presumes adiabatic conditions and uniform flow distribution at the filter entrance, laminar axial flow inside channels, pressure drop according to Darcy’s law for flow across the soot layer and substrate.
When dealing with pressure drop in [7] steady state, isothermal and non‐reacting conditions are assumed and because of that only the equations for mass and momentum balance are used. That is, the equations dealing with conservation of energy and the reaction rate are not included. In the work of Torregrosa [9], soot porosity was considered negligible and therefore the soot layer thickness was neglected compared to the wall thickness and because of that the use of constant velocity through the wall was allowed. But since then it has been proved that soot porosity can be in the order of 90 % under existing conditions, this means that the soot layer thickness variations can cause wall velocity differences in the flow direction. The new model in [7] has taken all of the above into consideration. An important underlying hypothesis for the above stated, is the assumption that the velocity through the porous wall is uniform along the length of the filter channel. This hypothesis was shown to be a realistic one. In [10], solutions for on‐board calculations are given but since they assume isothermal conditions there is a limited accuracy in the calculations because under normal transient operating conditions there are temperature gradients of the order of C. Of course there are simple solutions to this problem e.g. manufacturers could put two measure points in a DPF system and thereby calculate a good approximation of the mean filter temperature in a system but this is rather unlikely because of the high cost.
16
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Figure 8. Schematic of gas flow through the soot layer and filter wall. Figure and text from [10].
The total pressure drop:
(3)
is calculated with the assumption of uniform wall velocity. This is due to the flow friction in the channels. In a clean filter the contribution to the total pressure drop is more than 50 %. In heavily loaded filters this contributes substantially because the free‐flow channel diameter becomes very small. This is less important in low cpsi filters (100 cpsi (cells per square inch)/0.43 mm wall thickness) in both clean and loaded filters. starts from the Darcy pressure drop law and includes the inertial Forchheimer term. The mass flow rate is steady through the soot layer but the velocity varies due to changes in gas density and flow area and the gas density changes because of the local pressure. The permeability of the soot is actually reduced through the soot layer but the calculation can be simplified with the mean value of the inlet and outlet pressures. It was shown that this simplification introduced an error less than 1 %. This becomes dominant for soot loadings of more than 1 g soot/l filter. This contributes to more than 90 % when the soot loading is more than 3 g soot/l filter. This can be seen in figure 9. is described by Darcy’s law which describes the flow of a fluid through a porous medium with uniform wall velocity. In a clean filter its contribution to the total pressure drop is around 30 %. However, in low cpsi filters this is responsible for more than 60 % of the total pressure drop in the clean filter. and are of minor importance. But the contraction pressure drop increases as the soot loading increases due to the decrease in open front area.
17
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Figure 9. Computed pressure drop contributions as a function of soot loading: (a) 200 cpsi/0.38mm, (b) 100 cpsi/0.43mm. Fig and text from [7].
18
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
3. Theory and measurement techniques 3.1. Twoport The aim of measuring the sound pressure is to determine the acoustical two‐port for the DPF, i.e. the transfer matrix that describe how two state variables are affected by the DPF.
⎡ pu ⎤ ⎡T11 T12 ⎤ ⎡ pd ⎤ ⎢ q ⎥ = ⎢T T ⎥ ⎢ q ⎥ ⎣ u ⎦ ⎣ 21 22 ⎦ ⎣ d ⎦
(4)
where pu,p are the sound pressure and qu,p are the acoustical volume flow and u and p indicate upstream and downstream of the DPF. When the two‐port is known, the transmission loss can be calculated. Due to the reflections in the tube, the measured sound pressure will consist of both the incident sound pressure and the reflected sound pressure. To determine the acoustical volume flow the incident and reflected sound pressure has to be know separately according to the equations for acoustical volume flow. The sound pressure is simply the sum of the incident and reflected sound pressure.
( q$ = ( p
) − p )S / ρ c ,
p = p+ + p−
+
−
(5)
0 0
where p+ and p− are the incident and reflected sound pressures respectively, with a
eiωt dependence, S is the cross section area of the duct and ρ0 is the density and c0 is the speed of sound. To determine the both sound pressure the sound pressure at two positions must be know.
p1 = p + + p − p 2 = p + ei − k+ s + p − eik− s ,
(6)
where p1 and p2 are the sound pressure amplitudes at the two positions respectively with a
eiωt dependence. The wave numbers are given by
k± =
k0 , 1± M
19
(7)
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
where k0 = ω/c0 and M = Mach number [11]. The two sound pressures can be achieved in two different ways: either with one microphone which is moved between two positions or with two microphones flush mounted in the wall with a space, s, between them. The later method is called two‐microphone method, TMM, [12]. From equation (4) the incident and reflected sound pressure can be derived as,
( (− p e
p + = D −1 p1eik− s − p 2
p − = D −1
− ik+ s
1
)
)
+ p2 ,
(8)
(9)
where iMk0 s
⎛ k0 s ⎞ . 2 ⎟ ⎝ 1− M ⎠
−1 D = 2ie1− M sin ⎜ 2
Equations (8) is then inserted in equation (6) and then in equation (4). But since equation (4) has four unknowns and only two equations are given by equation (6), hence two more equations need to be found. This can be done by changing the condition of the system and there are two different methods: the two source method and the two load method. In the two source method the sound pressures are first measured with the loudspeaker at one side of the DPF and then with the loudspeaker at the other side of the DPF. In the two load method the boundary condition is changed by having two different terminations. But the different terminations must be independent at every frequency. This gives,
⎡⎛ pu ⎞ ⎛ pu ⎞ ⎤ ⎡T11 T12 ⎤ ⎡⎛ pd ⎞ ⎛ pd ⎞ ⎤ ⎢⎜ ⎟ ⎜ ⎟ ⎥ = ⎢ ⎥ ⎢⎜ ⎟ ⎜ ⎟ ⎥ ⎢⎣⎝ qu ⎠a ⎝ qu ⎠b ⎥⎦ ⎣T21 T22 ⎦ ⎢⎣⎝ qd ⎠a ⎝ qd ⎠b ⎥⎦ ,
(10)
where a and b denotes the measured sound pressure and particle velocity at the two different conditions. [11 and 13] It is convenient to rewrite equation (8) in terms of frequency response instead of amplitudes. This can be done by using the input signal, e$ , to the loudspeaker. −1
⎡T11 T12 ⎤ ⎡⎛ H p ,u ⎞ ⎛ H p ,u ⎞ ⎤ ⎡⎛ H p ,d ⎞ ⎛ H p ,d ⎞ ⎤ ⎢T T ⎥ = ⎢⎜ H ⎟ ⎜ H ⎟ ⎥ ⎢⎜ H ⎟ ⎜ H ⎟ ⎥ ⎣ 21 22 ⎦ ⎢⎣⎝ q ,u ⎠a ⎝ q ,u ⎠b ⎥⎦ ⎢⎣⎝ q ,d ⎠ a ⎝ q ,d ⎠b ⎥⎦
(11)
where H ji = j / e$ is the transfer function and j = p, v$ and i = u , d . 3.2. Microphone calibration If the TMM is used, the microphones must be phase calibrated. To perform the calibration all the microphones are exposed to the same sound pressure and then the transfer function can be calculated. If only two microphones are used the transfer function will be the one between
20
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
them, but if more than two microphones are used one of the microphones must be chosen to be a reference microphone which the others are compared to. 3.3. Pressure drop The pressure between two points (1 and 2) in a fluid can be described by Bernoulli’s equation (12). This is valid for gases at low Mach numbers at any arbitrary point along a streamline where the gravity is constant, where there is incompressible flow and therefore constant density,
p1 +
ρ u12 2
+ ρ gh1 = p2 +
ρ u22 2
+ ρ gh2 ,
(12)
where
p = static pressure (Pa)
ρ = density of the fluid (kg/m3) g = acceleration due to gravity (m/s2) u = fluid flow speed at a point on a streamline (m/s)
h = is the elevation of the point above a reference plane, with the positive z‐direction pointing upward, so in the direction opposite to the gravitational acceleration (m/s2). Because g, u, h and ρ are constant, what changes between the two points, is the static pressure Δp, and this is what is measured. This can be measured with a mini manometer.
3.4 Limitations Some limitations connected to this kind of measurement concern the frequency range. The theory behind the acoustical two‐port implies plane wave propagation which only is true for low frequencies. To ensure plane waves in the ducts the frequency range must, according to [12], satisfy
fu =
0, 58c0 , d
(13)
where fu is the upper limitation of the frequency, d the inner dimension of the duct and c0 the speed of sound at 20 °C. One of the limits for the lower frequency is the electrical noise which has a frequency of 50 Hz. Due to the fact that plane wave propagation is necessary for the method to be valid, another limitation is the distance between the microphones and the source of sound. According to [15] the wave propagation is, approximately, plane at a distance, r, from the source if,
r=
1
λ
,
where λ is the wave length. 21
(14)
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
According to SS‐En ISO 10534‐2, [12], the ducts must have rigid, smooth and non‐porous walls without holes or slits and the thickness of the wall must be thick enough so the ducts not start to vibrate due to the sound signal. For metal, circular ducts this implies a thickness of 5 % of the exterior diameter. Furthermore the microphones must be flush mounted in the walls and be of the identical type. More demands that can be applied in this kind of measurements can be seen in [12]. To avoid too large magnification of measurement errors, the frequency range must be limited according to,
0.1π ≤
k0 s ≤ 0.8π , 1− M 2
(15)
Equation (15) corresponds to a frequency range of three octave bands. If a larger frequency range is needed, more microphones must be used [13]. The models for the pipes in SIDLAB also assume plane wave propagation hence the frequency must be lower than the cut‐on frequency where modes of higher order will propagate. According to [14] the frequency must then satisfy,
f c = 1.84
c , πd
where fc is the cut‐on frequency and c the speed of sound.
22
(16)
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
4. Measuring The measurements were performed at room temperature (20°C) and 102 kPa in the test rig for measurements in ducts in MWL. A figure of the test rig can be seen in figure 10, and its dimensions can be seen in Appendix B. The transmission loss and the pressure drop were measured on a clean DPF and three with different soot layer thickness. The technical data for the filter and the soot can be seen in table 3 in Appendix A. The ducts in the test rig had the inner diameter of 90 mm and a wall thickness of 5 mm, hence the three times bigger DPF had to be placed in an expansion chamber with a conical inlet and outlet. The sound pressure was measured with three condenser microphones (Brüel & Kjær 1/4‐inch type 4938) and matching preamplifier (Brüel & Kjær, Falcon Range ¼ inch, type 2670), connected through a Microphone Conditioning Amplifier (Brüel & Kjær, Nexus model 2690‐A‐ OS4) to a SigLab interface. To be able to obtain the six microphone positions in the test rig, the sound pressure was first measured upstream of the DPF with three microphones and then downstream with the same three microphones. For each measurement (upstream and downstream) the location of the source was switched according to the two‐source method. The sources of sound were two loudspeakers connected to the duct. The measurements were performed with stepped sinus with five averages with the frequency range from 50 Hz to 1000 Hz, with the step of 10 Hz, as well as with random noise. This range satisfies the frequency range given from equation (15) (from 34 Hz to 1098 Hz with the microphone spacings, 125 mm and 495 mm). The transfer functions between the input and output signals were measured without flow for the DPF: s with a soot layer. This was done to protect the sensitive microphones from the soot. The clean filter was measured with and without flow (Mach 0.05 and Mach 0.1). The two‐port and the transmission loss were then calculated from the transfer functions with the program SIDLAB. The pressure drop and the flow speed were measured with a mini manometer (SwemaAir 300), a SWA‐07 transducer and a pitot tube for the flow speed measurements. The pressure drop was measured close to the DPF (see figure 10, were A and B indicate upstream and downstream), i.e. inside the expansion chamber but the flow speed was measured upstream the conical area expansion (see figure 10, marker C) with a pitot tube. To minimize the risk with soot particles in the air a ventilation hose pipe was connected to the test rig, downstream of the DPF. 23
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
24
Jesper Kristoffersson Linus Falk Lissel
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Acoustical absorbent material
Jesper Kristoffersson Linus Falk Lissel
Loudspeaker
Flow Acoustical absorbent material
Loudspeaker
Figure 10. Arrangement for measurements where 1‐6 indicate the microphone positions Before the measurements were done the microphones were phase calibrated with a calibrator designed and manufactured by laboratory engineer Danilo Prelevic at MWL, KTH. The microphones were placed in the same cross section in the duct. The end part of the duct that is not connected to the box containing the loudspeaker is a dead end with a reflecting plug. The loudspeaker was fed a signal containing random noise generated with SigLab. Microphone number 1 was designated to be the reference microphone.
25
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
5. Results The two‐port and transmission loss were calculated with the MATLAB program SIDLAB. It was only done for the complete system, i.e. altogether with the expansion chamber, this because of the complexity of the system and the lack of time. The results for the transmission loss without flow for the four DPF: s can be seen in figure (11).
Figure 11. The transmission loss for the filter including the expansion chamber. The transmission loss increases as the soot layer increases.
26
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Figure 12, show the transmission loss for the clean filter without flow and with a flow of 0.05 and 0.1
M.
Figure 12. The transmission loss for the clean filter at different flow speeds. The transmission loss increases only marginal as the flow speed increases.
The results of the pressure drop measurements were plotted in MATLAB and are presented in figure (13) and table 1. The pressure drop for the filters with soot loading was measured at a lower maximum flow speed, this to minimize the risk with loosening of soot particles. According to [1] the pressure drop can be written as
Δ p =R1U 0 + R2U 02
(18)
where U0 is the flow speed in the ducts. The coefficients in (18) are presented in table 2. They were calculated in a least square sense with the MATLAB command polyfit.
27
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
Figure 13. The measured pressure drop for the different filters. The pressure drop increase as the soot layer increase. Serial number: DY00480 Soot mass: 2 g soot/liter filter Flow speed 2.1 [m/s] Pressure drop 94
6.9
12.7
20
27
34
41
315
600
940
1310
1710
2140
5.5
11.4
17
24.5
32
39.3
340
670
1060
1500
1990
2500
4.5
8.7
14.5
20.4
26.5
33
370
760
1260
1850
2510
3230
Serial number: DY00481 Soot mass: 4 g soot/liter filter Flow speed 1.4 [m/s] Pressure drop 94
Serial number: DY00482 Soot mass: 10 g soot/liter filter Flow speed 1 [m/s] Pressure drop 108
Serial number: DY00843 Soot mass: no soot Flow speed [m/s] Pressure drop
1.6 14.7
5 52.5
12.5 150
20 240
25.5 28.5 377 410
Table 1. The pressure drop at different flow speeds for the filters.
28
30.7 510
43 750
53 970
62 1200
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
R1 [Ns/m3] R2 [Ns2/m4]
DY00480
DY00481
DY00482
DY00843
44.3 0.178
63.0 ‐0.028
91.8 0.046
9.85 0.170
Table 2. The coefficient for the pressure drop calculated in a least square manner in MATLAB.
A simulation for the expansion chamber system without the DPF was done in SIDLAB to be able to explain the reason behind the unusual transmission loss curve. This can be seen in figure 14.
Figure 14. The transmission loss, simulated in SIDLAB, for the expansion chamber with two conical inlets and two conical outlets. The dotted vertical line indicates the cut‐on frequency, 661.5 Hz, for the expansion chamber.
29
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
6. Discussion The aim of this project was to investigate how different soot layers in a DPF affect the transmission loss and the pressure drop over the filter. Due to that the filter dimension was three times bigger than the ducts in the test rig, an expansion chamber had to be built in which the DPF was to be placed. This expansion chamber made the system very complex and because of that it was a too difficult task to calculate the transmission loss and the two‐port only for the filter. Hence the results in this report concern the transmission loss and the two‐ port for the complete system, i.e. the filter including the expansion chamber. The expansion chamber and its dimensions are some of the reasons why our results have a different structure compared to the results of Åbom and Allam, which show a bent line describing the transmission loss, as opposed to the peaks and valleys as in our results. Due to the abnormal structure of the transmission loss, the measurements were also performed with random noise with 200 averages. This gave the same results as the measurement with step sinus; hence we considered the results to be correct. As can be seen in figure 15, which show the transmission loss simulated with SIDLAB for the expansion chamber without the filter, we can clearly see the resemblance with our results and thereby get some explanation for our measured results. Since the SIDLAB simulation is valid only for frequencies below the cut‐on frequency, i.e. 661.5 Hz for the system, this has to be considered when comparing the simulated and the measured results. Despite the unusual shape of the transmission loss curve, what can be seen is that the transmission loss increases as the soot layer increases, which was expected.
Figure 15. The transmission loss for the expansion chamber with two conical inlets and two conical outlets. The dotted vertical line indicates the cut on frequency, 661 Hz, for the expansion chamber.
As can be seen in figure 11, (nr. 1 in the results), the transmission loss for the clean filter with expansion chamber has a different structure than the others. The possible reason for this is
30
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
that the first minima in the transmission loss, at 80‐90 Hz, for the clean filter is canceled out in the soot loaded filters by the transmission loss due to soot layer. What can be seen in figure 12, (nr 2 in the results) is that the contribution to the transmission loss due to the flow is very small. This is concordant with the results of Åbom and Allam. Hence measuring the transmission loss and calculating the two‐port without flow still give accurate results. Regarding the pressure drop, this could be measured close to the filter, i.e. inside the expansion chamber, hence this results is valid for the filter itself. As expected it increases as the soot layer increases. The pressure drops for the clean DPF and the one with 4 g sot/l filter are not a smooth line. This is probably due to the fact that when doing the pressure drop measurements we started the fan at 80 % of its maximum power and then lowered the power with decrements of 10 % and then waited for the system to settle before we did the pressure drop and flow speed readings. We probably did not wait long enough for the system to settle before doing the measurements. But these errors are small and do not affect the coefficient in equations 18 too much when they are calculated in a least square sense for the flow speed range we used, which can be seen in figure 16.
Figure 16. The pressure drop for DPF: s. The circles are the measured pressure drop and the lines are calculated in a least square sense.
31
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
7. Conclusion The purpose of this project was to investigate how the transmission loss and the pressure drop over a diesel particulate filter (DPF) are affected by a soot layer. The conclusions after the measurements were:
the transmission loss, for the complete system, increases as the soot layer increases, the transmission loss are hardly affected by the flow in the ducts, hence doing the measurements without flow gives accurately results, the pressure drop increases as the soot layer increases, the SIDLAB‐simulation of the transmission loss for the system, is only valid up until 661 Hz, because that is where the cut‐on frequency for the system is located.
Further studies must be done to calculate the transmission loss and the two‐port for each DPF separately, i.e. without the expansion chamber.
32
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
8. References [1] S. Allam and M. Åbom. Acoustic modeling and testing of diesel particulate filters. Journal of Sound and Vibration 288. 2005, pages 255‐273 [2] S. Allam and M. Åbom. Sound propagation in an array of narrow porous channels with application to diesel particulate filters. Journal of Sound and Vibration 291. 2006, 882‐901 [3] S. Allam and M. Åbom. Modeling and Testing of After‐Treatment Devices. Journal of Sound and Vibration. vol. 128, 2006, pages 347‐356 [4] S. Allam. Acoustic Modeling and Testing of Advanced Exhaust System Components for Automotive Engines. Ph. D. Thesis Department of Aeronautical and Vehicle Engineering, KTH, Stockholm, 2004 [5] G. Wenzhi, F. Liming, N. Wenbo and W. Hui. Theoretical and experimental investigation on diesel particulate filters. Noise Control Engineering Journal. Vol. 56, Issue 4, 2008, pages 282‐ 287 [6] A.G. Konstandopoulos, J.H. Johnson. Wall‐flow diesel particulate filters‐their pressure drop and collection efficiency. SAE Technical paper. No. 890405, 1989 [7] A. Haralampous, I.P. Kandylas, G.C. Koltsakis and Z.C. Samaras. Diesel particulate filter pressure drop Part 1: modeling and experimental validation. Int.J. Engine Res. Vol.5 No.2, 2004 [8] Aj. Torregrosa, J.R. Serrano, F.J. Arnau, P. Piqueras. A fluid dynamic model for unsteady compressible flow in wall‐flow diesel particulate filters. Energy. Vol. 36 Issue 1, 2011, pages 671‐684 [9] E.J. Bissett. Mathematical model of the thermal regeneration of a wall‐flow monolith diesel particulate filter. Chem. Engng Sci., 1984, 39, 1233‐1244 [10] A. Haralampous, I.P. Kandylas, G.C. Koltsakis and Z.C. Samaras. Diesel particulate filter pressure drop Part 2: modeling and experimental validation. Int.J. Engine Res. Vol.5 No.2, 2004 [11] M. Åbom. Experimental investigation of a truck muffler. Course material SD2155 Flow Acoustics, KTH [12] SS‐En ISO 10534‐2, 1998, Acoustics – Determinations of sound absorption coefficient and impedance in impedances tubes – Part 2: Transfer‐function method. [13] M. Åbom. Measurement of the scattering‐matrix of acoustical two‐ports. Mechanical System and Signal Processing. 1991, 5(2), pages 89‐104 [14] M. Åbom, with others. SIDLAB Acousticals User Manual. Version 2.6. March 2011 33
SA104X: Examensarbete på Kandidatnivå i ljud, vibrationer och signaler
Jesper Kristoffersson Linus Falk Lissel
[15]H.P. Wallin, with others. (2001) Ljud och Vibrationer, (2nd Ed.) Stockholm: Norstedts Tryckeri AB, ISBN 91‐7170‐434‐5
34
Appendix A – Filter data The dimensions of the DPF: s in the references and the project. Filter name RC: 200/12 RC: 200/20 EX80: 100/17 EX80: 200/14
Channel width dh x 103 [m] 1.5 1.3 2.11 1.44
Wall Thickness ht x 104 [m] 3.04 5.08 4.3 3.55
Permeability σw x 1013 [m2] 25 25 2.5 2.5
‐5
n x 10 [channels/m2] 3.87 2.48 1.55 3.10
R1 [Ns/m3] 87.1 233.3 199.8 184.1
R2 [Ns2/m4] 29.2 41.56 30.92 39.2
Table 1. Data from the tested DPF units in ref. [1] including the linear (R1) and the quadric (R2) pressure drop (ΔP) coefficients (at 20°) determined from measurements: ΔP = R1 U+ R1 U. Table from [1].
Diameter/length [m] 150/250
n x 10‐5 [channels/m2] 3.10
Channel width dh x 103 [m] 1.44
Wall thickness ht x 104 [m] 3.55
Wall permeability σw x 1013 [m2] 2.50
Table2. Data for the studied DPF in ref. [1]. Table from [1].
Corning ID
Scania Part#
DY00480 DY00481 DY00482 DY00843
2004158‐1 2004159‐1 2004161‐1 2004162‐1
Soot load [g] 27.9 56.2 142.4 0
Cell Density [CPSI] 198 199 199 199
Wall Th [mils]//[mm] 11.8//0.30 12.4//0.315 12.5//0.318 12.2//0.31
Cement th [mm] 1.3 1.2 1.2 1.2
Plug length [mm] 10 10 10 10
Length [mm] 254.8 254.5 254.8 254.8
Diameter [mm] 266.7 266.7 266.7 266.7
Table3. Data for the studied DPF. The soot density = 2000 kg/m3 and the soot packing density ~ 100‐150 kg/m3. The soot permeability = 2.5‐4.5*10‐15 m2. A uniform soot distribution is assumed.
Appendix B – Dimension of the test rig
Appendix C – Datablad Bruel & Kjaer, Nexus model 2691AOS2 Signal Conditioner sn: 2301659 Charge Input CONNECTOR: TNC (TNC to 10.32UNF adaptor JP-0162 included) GROUNDING: Single-ended or floating MAX. INPUT: Differential Charge: 10 nC (peak); Type 2692-C/D: 100 nC (peak) Common Mode Voltage: 4.2V (peak) At gain ≥ 0.316 mV/pC (.10 dB gain with 1 nF transducer capacitance) INPUT PROTECTION: Differential Charge: ≤300 nC (peak) Common Mode Voltage: ≤15V (peak) COMMON MODE REJECTION RATIO: >50 dB (typical) (50 to 60Hz with 1 nF transducer capacitance) AMPLIFIER GAIN: 0.1mV/pC to 10 V/pC (.20 to +80 dB gain with 1 nF transducer capacity); Type 2692-C/D: 0.01 mV/pC to 10 V/pC (.40 to +80 dB gain with 1 nF transducer capacity) TRANSDUCER SENSITIVITY RANGE: 10.19 to 10.6 C/MU (MU = mechanical units: m/s2; g, N, lb., Pa) CALIBRATED OUTPUT: Selectable in 10 dB steps. 100 dB attenuator range, 10.15 to 107 V/MU (Type 2692-C/D: Selectable in 10 dB steps. 120 dB attenuator range, 10.16 to 107 V/MU) }1% for 0°C ≤ Ta ≤ +40°C and }2% for .10°C≤Ta ≤+55°C Frequency range from 5 × fl to 0.2 × fu fl = lower freq. limit: 0.1, 1.0 or 10 Hz fu = upper freq. limit: 0.1, 1, 3, 10, 30 or 100 kHz FREQUENCY RANGE (.10%): Acceleration: 0.1Hz to 100 kHz (transducer cable length <10m); Velocity (optional): 1.0Hz to 10 kHz Displacement (optional): 1.0Hz to 1 kHz LOW-PASS FILTER (.10%): 0.1, 1, 3, 10, 22.4, 30 or 100 kHz, attenuation slope 40 dB/decade HIGH-PASS FILTER (.10%): Acceleration: 0.1, 1.0 or 10 Hz; Type 2692-C/D: 0.1, 1 or 20 Hz Velocity (optional): 1.0 or 10 Hz Displacement (optional): 1.0 or 10 Hz INHERENT NOISE (2 Hz to 22.4 kHz): <5 fC referred to input, .10°C≤Ta ≤+40°C <10 fC referred to input, 40°C≤Ta ≤+55°C (amplifier sensitivity (>20 dB) with 1 nF transducer capacitance) HARMONIC DISTORTION AND NOISE (2 Hz to 22.4 kHz, Qin ≤2 nCpeak, Vout ≤3.16V peak): <0.003% for amplifier gain ≤0.1 V/pC (<40 dB gain with 1 nF transducer capacitance) ENVIRONMENTAL SUSCEPTIBILITY (referred to input): Magnetic Field: <0.2 fC/(A/m) Electromagnetic Field: <20 fC/(V/m) or <4 fC/V Vibration (10 to 500Hz): <30 fC/(m/s2) MOUNTED RESONANCE TESTING1: EP Patent 715.722, US Patent 5.753.793 Mounted resonance testing of the accelerometer and cable interconnection, controllable from front panel and RS.232 interface TEST TONE OSCILLATOR:
CHANNEL TO CHANNEL PHASE-MATCH: 2.1° . 0.1° × (f/fl) from fl to 20 × fl 0.1° from 20 × fl to 0.1 × fu (f/fu)° from 0.1(fu) to fu fl: lower freq. limit: 10 Hz fu: upper freq. limit: 0.1, 1, 3, 10, 30 or 100 kHz OPTIONAL FILTERS: Integration: Single and double Type 2692-C: Single and double integration filters included Type 2692-D: Single and double integration filters optional Microphone Input CONNECTOR: 7-pin LEMO GROUNDING: Outer shield grounded to chassis INPUT IMPEDANCE: 1MΩ|| 300 pF (AC coupled) MAX. INPUT: 31.6V (peak) INPUT PROTECTION: ≤50 V (peak) AMPLIFIER GAIN: .20 to +60 dB (80 dB with reduced specs.) TRANSDUCER SENSITIVITY RANGE: 10.12 to 103 V/MU (MU = mechanical units: Pa, mm) CALIBRATED OUTPUT: Selectable in 10dB steps. 100 dB attenuator range, 10.15 to 107V/MU ( }0.1 dB for 0°C ≤ Ta ≤ +40°C and }0.2 dB for .10°C≤Ta ≤+55°C Frequency range from 5 × fl to 0.2 × fu fl = lower freq. limit: 0.1 or 20 Hz fu = upper freq. limit: 0.1, 1, 3, 10, 22.4, 30 or 100 kHz) POLARIZATION VOLTAGE ( }0.5V or }0.25%): 0 or 200 V (all channels simultaneously selected, short-circuit protected) PREAMPLIFIER SUPPLY: Fixed }14 V, }40 V or controlled automatically in accordance with input range (short-circuit protected) FREQUENCY RANGE (.1 dB): 0.1Hz to 100 kHz (gain ≤60 dB) (complies with IEC 1260 Class 0 and ANSI S1.11 Type 0.AA for fl = 0.1Hz and fu = 100 kHz) HIGH-PASS FILTER (.1 dB): 0.1Hz, attenuation slope 40 dB/decade or 20 Hz, attenuation slope 80 dB/decade LOW-PASS FILTER (.1 dB): 0.1, 1, 3, 10, 22.4, 30 or 100 kHz, attenuation slope 40 dB/decade A-WEIGHTING FILTER: Complies with IEC 61672.1 INHERENT NOISE (referred to input, gain >20 dB): <2 μV A-weighted HARMONIC DISTORTION AND NOISE (2Hz to 22.4 kHz, Vin ≤ 20V peak, Vout ≤ 3.16V peak): <0.003% for amplifier gain ≤40 dB ENVIRONMENTAL SUSCEPTIBILITY (referred to input): Magnetic Field: <0.2 μV/(A/m) Electromagnetic Field: <10 μV/(V/m) or <10 μV/V Vibration (10 to 500 Hz): <2 μV/(m/s2) CHARGE INJECTION CALIBRATION1: Verification of the entire measurement set-up including the microphone, preamplifier and connecting cable. Controllable from front panel and RS.232 interface
Spe cific atio ns
ω= 1000 rad/s (159.2Hz), sinusoidal Test Level: 1mV to 10 V ( }1%). Controllable from front panel and RS.232 interface Reference Tone: 1V (RMS), ( }1%), 159.2Hz RISE TIME: >7.5V/μs (100 kHz bandwidth)
Reference Tone: 1V (RMS) }1%, 1 kHz OVERLOAD DETECTION: Microphone preamplifier overload detection with respect to cable length (3 to 1000 meter) RISE TIME: >7.5V/μs (100 kHz bandwidth) 1. Brüel& Kjær patent CHANNEL TO CHANNEL PHASE-MATCH:
5.1° . 0.1° × (f/fl) from fl to 50 × fl (fl=0.1Hz) 2.1° . 0.1° × (f/fl) from fl to 20 × fl (fl=20Hz) 0.1° from 50× fl to 0.1 × fu (fl = 0.1Hz) 0.1° from 20 × fl to 0.1 × fu (fl = 20 Hz) (f/fu)° from 0.1(fu) to fu Where: fu = upper frequency limit: 0.1, 1, 3, 10, 22.4, 30 or 100 kHz OPTIONAL FILTERS: A-, B-, C- and D-weighting (one module) (complies with IEC 61672.1) Intensity Input Specifications as for microphone input, except when using the .Intensity. filter. CONNECTOR: 7-pin LEMO (two connectors on two input modules . adaptor required) CHANNEL TO CHANNEL PHASE-MATCH AND FREQUENCY RESPONSE (with .Intensity. filter (20Hz HP/22.4 kHz LP, 40 dB/ decade): Complies with IEC 61043 standard Class 1 and ANSI S1.9.1996 Class 1, with Brüel & Kjær sound intensity probes. (Conditions: output sensitivity for the two channels must be equal. Transducer sensitivity must be equal within 0.5 dB) DeltaTron® Input CONNECTOR: BNT GROUNDING: Single-ended or floating INPUT IMPEDANCE: 1MΩ || 100 pF (AC coupled) MAX. INPUT: Differential Voltage: ≤31.6V (peak) Common Mode Voltage: 4.2V (peak) INPUT PROTECTION: Differential Voltage: ≤50 V (peak) Common Mode Voltage: ≤15 V (peak) COMMON MODE REJECTION RATIO: 50 dB (50 to 60 Hz) (typical) AMPLIFIER GAIN: .20 to +60 dB gain (80 dB with reduced specs.) TRANSDUCER SENSITIVITY RANGE: 10.12 to 103 V/MU (MU = mechanical units: m/s2, m/s, g, N, lb., Pa) CALIBRATED OUTPUT Selectable in 10 dB steps. 100 dB attenuator range, 10.16 to 107 V/ MU. ( }0.1 dB for 0°C ≤ Ta ≤ +40°C and }0.2 dB for .10°C≤Ta ≤+55°C Frequency range from 5 × fl to 0.2 × fu fl = lower frequency limit: 0.1, 1.0 or 10 Hz fu = upper frequency limit: 0.1, 1, 3, 10, 22.4, 30 or 100 kHz) CONSTANT CURRENT SUPPLY ( }15%): +4mA or +10mA with a +28V voltage source Tacho Probe Supply: +8 V DC at BNT inner shield (short-circuit protected) FREQUENCY RANGE (.10%): 0.1Hz to 100 kHz (gain <60 dB) attenuation slope 40 dB HIGH-PASS FILTER (.10%): 0.1Hz or 1.0Hz (with attenuation slope 40 dB) or 10 Hz (with attenuation slope 60 dB/decade) LOW-PASS FILTER (.10%): 0.1, 1, 3, 10, 22.4, 30 or 100 kHz INHERENT NOISE (referred to input, gain >20 dB): <2.4 μV A-weighted <3.3 μV lin. 2Hz to 22.4kHz HARMONIC DISTORTION AND NOISE (2 Hz to 22.4 kHz, Vin ≤ 20V peak, Vout ≤ 3.16V peak):
<0.003% for amplifier gain ≤40 dB RISE TIME: >7.5V/μs (100 kHz bandwidth) ENVIRONMENTAL SUSCEPTIBILITY (referred to input): Magnetic Field: <0.2 μV/(A/m) Electromagnetic Field: <3 μV/(V/m) or <3 μV/V Vibration (10 to 500 Hz): <2 μV/(m/s2) OVERLOAD DETECTION: Preamplifier overload detection with respect to cable length (3 to 1000 meter) CHANNEL TO CHANNEL PHASE-MATCH: 5.1° . 0.1° × (f/fl) from fl to 50 × fl (fl=0.1 or 1Hz) 2.1° . 0.1° × (f/fl) from fl to 20 × fl (fl=10Hz) 0.1° from 50 × fl to 0.1 × fu for fl = 0.1, 1Hz 0.1° from 20 × fl to 0.1 × fu for fl = 10 Hz (f/fu)° from 0.1(fu) to fu Where: fu: upper freq. limit: 1, 3, 10, 22.4, 30 or 100 kHz Reference Tone: 1 V (RMS) }1% (0.1 dB), 1 kHz OPTIONAL FILTERS: A-, B-, C- and D-weighting (one module) (complies with IEC 61672.1) Integration: single and double (one module) Other filters available upon request General Specifications Power Supply INTERNAL BATTERY (not included): Nickel-Metal Hydride rechargeable battery supporting SMBus and on-battery charge level meter. Provides typically 15 hours of continuous use with a single channel and 4 hours with four channels without backlighting and without optional filters. With backlighting on, and with optional filters, battery provides typically 3 hours of continuous use. If NEXUS is not used for more than a month, please remove the battery to prevent discharging. Charging time is approximately 4 hours EXTERNAL DC POWER INPUT: Complies with ISO 7637.1 (12 V) and 7637.2 (24V) Input Range: 10 to 33VDC MAINS SUPPLY: Supported via supplied Mains Adaptor ZG-0426 (included), 90 .264VAC, 40 . 65Hz Digital Control Interface SERIAL INTERFACE: Conforms to EIA/TIA-574 (RS.232) Baud rate: 2400, 4800, 9600 Parity: None Data Bits: 8 Stop Bits: 1 Handshake: X-on/X-off .Plug and play. interface coupling Communication speed for a baud rate of 9600: Transmission time for one command of 5 characters is ~ 4ms Transmission time for one command of 5 characters and to receive an echo after each character is ~ 8ms. Execution time for one command is 100 ms to several seconds. Time to configure a complete 4-channel NEXUS using short form set-up with approx. 600 characters requires transmission time of 2 to 3 s (4 to 6 s with echo after each character). Execution time in NEXUS is from 40 to 60 s. For set-ups with over 1000 characters, the transmission time will be increased by at least 30 s due to delay in emptying receiver buffer. Response time after requesting a status of one load: < 0.5 s Response time after requesting a peak meter reading: < 0.5 s
Display Interface DISPLAY: 64 × 128 pixel graphical display with back-lighting on/off OVERLOAD DETECTION: On both common-mode and differential signals applied before filters. LED overload indication at the front panel and overload indication via RS.232 control interface Peak Meter DYNAMIC RANGE: .30 to +10 dBV (peak) RESOLUTION: 1 dB Analogue Output CONNECTOR: BNC GROUNDING: Single-ended or floating OUTPUT IMPEDANCE: = 50Ω//500 pF MAX. OUTPUT (differential voltage): 3.16V peak (6.32 V peak to peak); Type 2692-C/D: 10 V peak (20V peak to peak) MAX. DC OFFSET: } 25mV, typically < 2mV OUTPUT PROTECTION: Differential Voltage: ≤50V(peak) Common Mode Voltage: ≤15 (peak) Common Mode Rejection: >50 dB (50 to 60 Hz) for Common Mode Voltage ≤2V peak (voltage injected into instrument) OUTPUT DRIVE CAPACITY: 100m of cable length (100 pF/m) to 20 kHz 1000 m of cable length (100 pF/m) to 2 kHz CHANNEL SEPARATION: better than . 100 dB at 1 kHz
Dimensions and Weight Height: 90 mm (3.5″) Width: 144 mm (5.7″) Depth: 230 mm (9.1″) Weight: Approx. 3 kg (6.6 lb.), for a 4-channel unit including battery Note: All values are typical at 25°C (77°F), unless measurement uncertainty is specified. All uncertainty values are specified at 2σ (i.e. expanded uncertainty using a coverage factor of 2) Calibration NEXUS amplifiers are supplied with a Manufacturer.s Certificate of Conformance. An initial calibration can be supplied as an option. The calibration services in the table below are based on one channel. There is an extra charge for each additional channel: All other accessories are listed in the associated ordering information sheet (BA 0287).
Appendix D Microphone Preamplifier Type 2670 Specifications – Falcon Range 1/4-inch Microphone Preamplifier Type 2670 Brüel & Kjaer FREQUENCY RESPONSE (re 1 kHz) 15 Hz to 200 kHz, ±0.5 dB. See Fig. 1
MAX. OUTPUT CURRENT 20 mA (peak) Note: The max. output current can be limited by the power supply
ATTENUATION 0.4 dB (max.)
MAX. OUTPUT VOLTAGE Maximum output voltage Vp-p is equal to total supply voltage minus 10 V
PHASE LINEARITY ≤±1° from 20 Hz to 100 kHz (20 pF mic. capacitance)
DISTORTION (THD) Less than –80 dB at 25 V out, 1 kHz
PHASE MATCHING 0.015° at 50 Hz (20 pF mic. capacitance)
NOISE 14.0 ∝V Lin. 20 Hz – 300 kHz (typical) 30.0 ∝V Lin. 20 Hz – 300 kHz (max.) 4.0 ∝V A-weighted (typical) 7.0 ∝V A-weighted (max.)
OUTPUT SLEW RATE 2 V/∝s INPUT IMPEDANCE 15 GΩ || 0.25 pF (typical)
POWER SUPPLY, DUAL ±14 V to ±60 V
OUTPUT IMPEDANCE 50 Ω (max.)
POWER SUPPLY, SINGLE 28 V to 120 V
CONNECTOR TYPE LEMO type FGG.1B.307
OUTPUT DC OFFSET ≈1 V for a dual supply, or 1
PIN CONNECTIONS
≈ /2 the voltage of a single supply
CURRENT CONSUMPTION 3 mA plus1 output current
6 5
7 4
Cable's output plug seen from outside
CALIBRATION INPUT 2 Charge insert capacity, typically 0.2 pF Max. 10 V RMS, input capacitance: 1 nF 3
DIMENSIONS Diameter: 6.35 mm (0.25″) Length: 89.5 mm (3.5″)
950236e
MAX. CABLE RELEASE 5 kg Pin 1
Calibration input
2
Signal ground
3
Polarization voltage
4
Signal output
5
Not connected
6
Power supply positive
7
Power supply negative/ground
Casing
Optional Accessories LEMO to 7-pin Brüel & Kjær adaptor Adaptor (1/4- to 1/8-inch) Microphone Handbook Insert Junction Adaptor for Charge-injection Calibration
Local representatives and service organisations worldwide
Charge Injection Calibration Technique patented according to US Patent No. 5,400,297
Connected to instrument chassis
Type 2670 1/4-inch Microphone Preamplifier (2 m fixed cable with LEMO connector). Supplied in a robust plastic box.
ZG 0350 UA 0160 BA 5105 WB 0850
LEMO Note: The above are valid for the fixed 2 metre cable and 6.2 pF microphone capacitance unless otherwise specified.
Extension Cables LEMO to LEMO: AO 0414 3 m (9.8 ft.) AO 0415 10m (32.8 ft.) AO 0416 30m (98.4 ft.) EL 4004/xx Length xx m (specified by customer) AR 0014 Flat Cable
Appendix E – Datablad Microphone ¼″ Pressurefield Microphone — Type 4938, sn: 2323070, 2491541, 2239162, Brüel & Kjaer Type 4938 is designed for high-level and highfrequency measurements where a microphone with pressure-field response is needed. Being externally polarized, Type 4938 must be used with a classical preamplifier. USES • High-level measurements • High-frequency measurements • Flush measurements • Measurements at high temperatures
FEATURES • Sensitivity: 1.6 mV/Pa • Frequency: 4 Hz – 70 kHz • Dynamic Range: 30 – 172 dB –40 to +300°C (–40 to +572°F) • Temperature: • Polarization: 200 V External
Use of Pressure-field Microphones
Polarization Voltage
Pressure-field microphones are optimised to have a flat frequency response in a pressure field. Such microphones should be used in small, closed couplers or close to hard, reflective surfaces. For optimum performance at higher frequencies or when flush mounted in an airflow, Type 4938 should be used without its protection grid.
Type 4938 requires an external polarization voltage and must therefore be used with a classical preamplifier. Externally polarized microphones may be used at higher temperatures without severe changes in sensitivity.
• Type 4938 is suited for all high-level, highfrequency measurements where a robust and stable pressure-field microphone is required • Type 4938-W-001 is a variant optimised for use in high static pressure • Type 4938 with WB-1418 is for measurement of airbag deployment noise according to SAE J247, Feb., 1987
TEDS Microphones Type 4938 is available in TEDS combinations with different classical preamplifiers. The TEDS microphone is considered as a single unit and has been sealed in a clean environment. The TEDS is programmed with the loaded sensitivity of the actual cartridge and the data is therefore readily available in the transducer. The default TEDS template is according to IEEE P1451.4 but TEDS to IEEE 1451.4 is available on request.
Manufacturing and Stability A laser-welded, stainless-steel diaphragm ensures superior long- term stability and mechanical robustness – Type 4938 will withstand the 1 m drop test of IEC 60068–2–32.
The design of Type 4938 makes it especially suited for flush mounting. All Brüel & Kjær Measuring Microphones are assembled in a clean room. This ensures that the microphones maintain their inherent noise floor and high stability, even when used in environments with a combination of high humidity and high temperature.
Individual Calibration Data Each Type 4938 comes with an individual calibration chart including information about the open-circuit sensitivity and the pressure-field frequency response. An enclosed mini-CD contains the individual calibration data at 1/12- octave frequencies plus a wealth of technical information, such as the influence of different accessories, corrections in different sound fields and much more. Using the CD data and the REq-X feature of PULSE™, a real-time correction for different measurement situations, can increase measurement accuracy.
Fig. 1 Typical pressure response of Type 4938 without protection grid. The low-frequency response is valid when the vent is exposed to the sound field dB 5 0 -5 -10 -15 -20 1
10
100
1000
10000
100000 Hz 080111
Specifications – ¼″ Pressure-field Microphone Type 4938 (valid from serial number 2496492) IEC 61094–4 Type Designation: WS 3P Polarization Voltage: 200 V Opena circuit Sensitivity (250 Hz) : −56 ±3 dB re 1 V/Pa, 1.6 b mV/Pa 0.4 mV/Pa ±3 dB a Pressure-field Response : 4 Hz to 70 kHz: ± 2 dB a Lower Limiting Frequency (−3 dB) : 0.3 to 3 Hz Lower Limiting Frequency with Modification WB-1418: <0.15 Hz Pressure Equalization Vent: Side vented
Diaphragm Resonance Frequency: 60 kHz (90° phase shift) a Polarized Cartridge Capacitance : 6.1 pF (250 Hz) 3 Equivalent Air Volume: 0.25 mm (250 c Hz) Cartridge Thermal Noise : 30 dB(A), 37 dB(Lin., 20 Hz to 100 kHz)
Upper Limit of Dynamic Range d (3% Distortion): >172 dB SPL
Vibration Sensitivity (< 1000 Hz): 69 dB
Max. Sound Pressure Level: 182 dB (peak)
Magnetic Field Sensitivity: 10 dB SPL for 80 A/m, 50 Hz field Estimated Long-term Stability: > 1000 years/dB (at 20°C (68°F)) > 100 hours/dB (150°C (302°F))
ENVIRONMENTAL Operating Temperature Range: e −40 to +150°C (−40 to +302°F) Storage Temperature: In Microphone Box: −30 to +70°C (− 22 to +158°F) With Mini-CD: 5 to 50°C (41 to 122°F) Temperature Coefficient (250 Hz): 0.003dB/K (−10 to +50°C, 14 to 122°F) Pressure Coefficient: − 0.003 dB/kPa Operating Humidity Range: 0 to 100% RH (without condensation) Influence of Humidity: < 0.1 dB in the absence of condensation d. 170 dB (peak) with ±15 V preamplifier supply voltage. Airbag noise set >177 dB SPL peak
a. Individually calibrated b. With WB-1418, and Type 2670 with WB-1419 c. Not valid with modification WB-1419
e. Can be used up to 300°C (572°F) but with a permanent sensitivity change of typ. 0.6 dB which stabilises after 1 hour
Type 4938 ¼″ Pressure-field Microphone 4938-W-001 Optimised for high static pressure 4938/WB-1418 For airbag deployment noise Include the following accessories:
4938-C-002 ¼″ Pressure-field Microphone, with Preamplifier 2669-C with Adaptor UA-0035 4938-L-002 ¼″ Pressure-field Microphone, with Preamplifier 2669-L with Adaptor UA-0035
• BC-0229: Calibration Chart
f
f • BC-5002: Microphone Mini-CD
TEDS COMBINATIONS 4938-A-011 ¼″ Pressure-field Microphone, with Preamplifier 2670 4938-B-002 ¼″ Pressure-field Microphone, with Preamplifier 2669-B with Adaptor UA-0035 f. Quote microphone serial number if re-ordering calibration data
Local representatives and service organisations worldwide
OPTIONAL ACCESSORIES Type 2670 ¼″ Microphone Preamplifier Type 4231 Sound Calibrator Type 4228 Pistonphone Type 4226 Multifunction Acoustic Calibrator DP-0775 Calibration Adaptor for ¼″ Microphones UA-0033 Electrostatic Actuator DB-0264 ¼″ Adaptor for UA-0033
2
equivalent SPL for 1m/s axial vibration
DIMENSIONS Diameter with Grid: 7 mm (0.27″) Diameter without Grid: 6.35 mm (0.25″) Height with Grid: 10.5 mm (0.41″) Height without Grid: 9 mm (0.35″)
Thread for Preamplifier Mounting: 5.7 mm–60 UNS Note: All values are typical at 23°C (73.4°F), 101.3 kPa and 50% RH unless otherwise specified Compliance with EMC Directive
UA-0035 WA-0371 UA-0122 UA-0123 EU-4000 UA-0385 WQ-1099 WQ-1133
¼″ to ½″ Adaptor, L = 72.5 mm ¼″ to ½″ Adaptor, L = 32 mm Flush Mounting Kit (angle) Flush Mounting Kit (straight) ¼″ Angle Piece Nose Cone for ¼″ Microphone Windscreen for ¼″ Microphone, 65 mm diameter Windscreen for ¼″ Microphone, ellipsoidal
CALIBRATION SERVICES 4938-CAI Accredited Initial Calibration 4938-CAF Accredited Calibration 4938-CFF Factory Standard Calibration
Appendix F – Calibrator The calibrator is built by Danilo Prelevic, MWL, KTH and is constructed in such a way that holders for the microphones are placed in the same cross section in a tube. In that way the membrane of the microphones are placed in the same cross section in the tube and they are then able to measure the same sound pressure in the plane wave area of the tube. The end section after the microphones is a dead end and the other end is mounted in a box containing a speaker. The speaker is fed with a white noise signal, generated by SigLab.
Figur ?? Calibration head
Figur 1 Calibration speaker and tube
Appendix G Datablad SwemaAir 300 Generellt Noggrannheten för varje givare är noggrannheten tillsammans med SwemaAir 300. Noggrannheten för varje givare innehåller alltså noggrannheten för SwemaAir 300, utbytbarheten, kalibreringen, temperaturberoendet, hysteresen och repeterbarheten. Användaren ska korrigera mätningarna med korrigeringarna i kalibreringsbeviset för att uppnå uppgiven noggrannhet. Utprovad enligt: EN 50081‐1 och EN 50082‐1 SwemaAir 300, sn: 389739 Arbetsområde SwemaAir 300 0...+50°C Minneskapacitet, Strömoberoende: anteckningsbok ‐ loggbok ca 200 protokoll eller ca 1500 mätpar, 2500 enskilda mätvärden. Utbytbarhetsnoggrannhet Försumbar i förhållande till givarnas noggrannhet. Analog utsignal Se varje givare. Digital utsignal RS232, skicka : så ger instrumentet displayens (stora siffror) mätvärde. (max överföring: 5 gånger per sekund) Batteri 9V IEC 6 F22 Batterilivslängd Alkaliska batterier: 6...10 timmar, beroende på givare och lufthastighet. Genom att trycka MAX * MODE samtidigt och i den ordningen, med en givare ansluten och därefter CLEAR visas mjukvaruversio med stora siffror och hårdvauversion med små siffror.
Appendix H Tryckgivare, SWA 07, sn:372909 Mätområde tryck Max belastning Noggrannhet Lägst
± 7 000 Pa ± 35 000 Pa ±1Pa ± 2% avläst värde vid 20°C Efter nollställning: ± 0,3 Pa ± 2% avläst värde 0,4 Pa/°C 0…50°C
0,1 %FSO/år 0,1 %FSO/år
Torr och fuktig luft, icke aggressiva gaser 1 mV/Pa, 100 mV/m/s Uppdateras 4ggr/s vid tidskonstant 0,25s Rekommenderas var 12:e månad.
Tempberoende Arbetstemperatur Långtidsstabilitet Nollpunkt Fullt mätområde Mätmedia Utsignal Kalibrering
Appendix I Datablad NAD C370, SR: H0ZC37006007
• 120W x 2 Continuous Power (8 ohms) • 210W, 340W, 450W Dynamic Power into 8, 4 and 2 ohms, respectively • Bridgeable to 300W mono (8 ohms) • “Building Block Concept”: Power amp section identical to C270 • A+B speaker switching on front panel & remote control • Impedance Sensing Circuitry (ISC) 12V trigger output for Stand-By/On switching of remote components Full System Remote control including Stand-by/Off • Relay Input Switching • Toroidal Power transformer • 7 Line inputs, including two tape in/outs • All discrete circuitry • Short signal path from input to output Tone controls defeat switch • 2 pre-amp outputs of which one is variable from 0dB to -12dB • Main-in • Soft ClippingTM • NAD Link After so many illustrious and world-wide acclaimed predecessors, NAD is proud to up the stakes yet again by introducing the all new C370, which in the best NAD tradition, offers a perfect blend of performance, flexibility and ease of use at a modest price. Features: The C370 is fully remote controlled and comes supplied with the NAD C Series system remote control. As the C370 has NAD Link, the remote control will also operate many other NAD products such as CD players, tuners, etc. Flexibility is another NAD strong point. The C370 has 7 line inputs (including 2 tape in/outputs with dubbing facility) and the pre-amplifier section can be separated from the power amplifier for easy system upgrades or adding ancillary equipment. The C370 sports 2 pre-amp outputs, and one of them is vari-able from 0dB to -12dB. Many systems benefit from the use of multiple power amplifiers for "Bi-Amping" (using separate power amplifiers to drive the bass and treble section of a loudspeaker) but not all power amplifiers are identical in gain (amplification factor). With the Variable Pre-out 2 facility differences in loudness between the C370's own power amplifier section and an external power amplifier can be dialled out precisely.
Two pairs of speakers can be connected and individually switched on or off (either via remote control or front panel), this could be a second pair in the listening room but also an additional set of speakers in, for example, the kitchen.
As the built-in power amp of the C370 is identical to that of the NAD Model C270 you can easily nearly triple the output power to your speakers by adding a C270 to your system. Bridge both power amplifiers and you have increased the power of your system to 2 x 300 Watts Continuous Power into 8 ohms. Thanks to this "Building Block Concept" your C370
need never become obsolete as system improvements are made. For remote on/off switching of ancillary components in a system, such as power amplifiers or active speakers, the C370 is equipped with a 12V trigger system. When switching the amplifier on, the 12V trigger output is also activat-ed which in turn can activate a 12V trigger input and switch on the remote devices, such as a C270 power amplifier, for example. Besides the 12V trigger, the C370 also has an AC switched outlet so you can easily switch your entire system on or off with the remote control or on the front panel. It is fashionable to omit tone controls nowadays: However, provided that the tone controls are properly designed, they can be really useful tools in making improvements to the overall sound. The C370 tone controls only work at the frequency extremes, leaving the critical mid-band essentially unaltered. The tone control circuits can be completely bypassed by using the tone defeat switch.
The C370 also incorporates NAD's acclaimed switchable "Soft Clipping" cir-cuit, which significantly reduces the risk of damage to loudspeakers due to prolonged high power operation. Impedance Sensing Circuit (ISC):
The C370 also benefits from NAD's proprietary Impedance Sensing Circuit (ISC) topology, now well established and used in many NAD models including the highly reviewed C320 and C340. The ISC topology allows the C370 to deliver maximum performance under virtually any circumstance, independent of the loudspeakers it is driving. The circuitry automatically senses the impedance characteristics of the loudspeak-er and will then adjust its power supply settings to best cope with that specific load. With ISC topology, the relationship of voltage to current is kept at an ideal proportion, eliminating the current limiting distortion common in lesser designs. This also gives it an unusual characteristic compared to traditional amplifiers when measuring its continuous output power; the RMS output
power remains the same at 120 Watts with either an eight or a four ohms load. This is not unusual for NAD however. NAD takes a stance to the mindless "brochure power" approach which does-n't give a realistic indication of an amplifier's true capabilities. Instead, the ISC topology is a practical approach to enable an amplifier to easily deal with musical dynamics and difficult speaker loads. More meaningful in the real world are the C370's dynamic capabilities; up to 450 Watts into 2 ohms and up to 70 amps current capability into 1 ohm!
Other design features include relay input switching (located right behind the input sockets for short signal paths) and the large Toroidal transformer (less mechanical hum and stray magnetic field). The generous power supply and over-specified output devices further lead to low noise and low distortion for the entire amplifier. Like all the other input and output sockets, the Preout/Main-in sockets are gold plated for long-term contact reliability. The headphone socket will drive virtually any non-electrostatic headphone.
PROVISIONAL SPECIFICATIONS - NAD C370 Continuous Average Power Output into 8 Ω
Power Amp Section
120W (20.8dBW)
(Min. power per channel, 20Hz - 20kHz both channels driven with no more than rated distortion) Rated distortion (THD 20Hz-20kHz)) Clipping power (max. continuous power per channel)
IHF dynamic headroom at 8 Ω IHF dynamic power
8Ω 4Ω 2Ω (ref. 8 Ω , 50Hz) R and C ref. rated power 1.1V
Damping factor Input impedance Input sensitivity V oltage gain Frequency response
20Hz-20kHz 3Hz/70kHz A weighted; ref. 1W A weighted; ref. rated power
Signal/noise ratio
0.03% 140W 2.4dB 210W (23.2dBW) 340W (25.3dBW) 450W (26.5dBW) >150 20kΩ / 300pF 23dB +/-0.3dB -3dB >100dB >120dB
Overall. Line in, Speaker out
Sensitivity; max vol. re. rated power Signal/noise ratio
300mV (24.8dBW) A weighted; ref. 1W >94dB A weighted; ref. rated power >115dB <0.03%
THD from 0.25W to rated power Continuous average power output into 8 Ω IHF dynamic headroom at 8 Ω IHF dynamic power
Bridged Mode
8Ω 4Ω
300W (24.8dBW) 3dB 600W (28dBW) 800W (29dBW)
Pre-amp Section Line Level Inputs
300kΩ / 320pF R and C ref. rated power 290mV 20Hz-20kHz +/-0.2dB
Input impedance Input sensitivity Frequency response Line Level Outputs
75 Ω Z Source + 1kΩ 220 Ω >100dB Yes
Pre-amp fixed output Tape Phones Signal/noise ratio Remote control Physical Specifications
Dimensions (W x H x D) Net weight Shipping weight
17 1/8 x 5 1/4 x 13 13/16"
(435 x 132 x 350mm) 26.45 lbs (12kg) 31.65 lbs (14.36kg)
Ganttschema
Arbetes gång Arbetet börjad med att vi, måndagen den 17 januari, fick projektuppgiften. Gantt‐schemat ovan gjordes omgående för att få en överblick av projektet och tidsåtgången för varje moment. Vårt första möte med Tony Karlsson på Scania bokades till den 31 januari. Vi bestämde oss för att ha gjort större delen av informationssökningen innan mötet, vilket vi även lyckades med. Dock han vi inte läsa igenom artiklarna innan möte. Första mötet med Mats Åbom hade vi den 1 februari och vi bestämde att vi skulle se ungefär en gång i veckan/varannan vecka. Våra möten blev dock lite mer sällan än så. Tiden för att läsa artiklarna satte vi till två veckor och en vecka för att sammanställa dem. Detta visade sig dock vara lite väl optimistiskt; tiden det tog att läsa dem var betydligt längre, lika så sammanställningen. Det var inte förrän dagarna innan redovisningen av informationssökningen den 25 mars som vi var klara. Efter redovisningen justerades en del efter feedback från redovisningen. Tanken var från början att vi skulle göra mätningarna vecka 14 för att ha god tid att analysera resultaten och kunna skriva rapporten. Så blev det tyvärr inte. På grund av strul med kopplingsanordningen blev mätningarna förskjutna drygt en vecka, d.v.s. till slutet av vecka 15 och början av vecka 16. Samtidigt blev datumet för slutredovisningen tidigarelagt ett par dagar vilket innebar att tiden för analys och rapportskrivande krympte. För att spara lite tid började vi på rapporten innan mätningarna påbörjades. Att analysera resultaten i SIDLAB visade sig vare lättare sagt än gjort. Vi var tvungna att installera programmet flera gånger innan vi fick det att fungera. Detta tog mycket värdefull tid. Inför provredovisningen och slutredovisningen på Scania var rapporten i princip klar.
Arbetsfördelning Arbetet med detta kandidatjobb har förflutit förvånansvärt smidigt, utan några gräl eller meningsskiljaktigheter. Vi har i stort sett endast arbetat tillsammans under alla dessa veckor, om man bortser från att Linus alltid lyckas komma ungefär 20‐30 min innan Jesper VARJE morgon. Även om vi i innehållsförteckningen nedan angett ”vem som gjort vad”, har vi båda två varit delaktiga i varandras skrivande, funderande och slutsatser. Abstract ‐ Jesper ............................................................................................................................ 2 Sammanfattning ‐ Jesper............................................................................................................... 3 Acknowledgements ‐ Jesper.......................................................................................................... 4 Table of contents ‐ Linus ............................................................................................................... 5 1. Introduction ‐ Jesper ................................................................................................................. 7 2. Collection of information ‐ Intro Jesper .................................................................................... 8
2.1. Acoustical models ‐ Linus................................................................................................ 8
2.1.1. Measurement technique ‐ Linus .......................................................................... 8
2.1.2. The models ‐ Linus ............................................................................................... 9
2.1.3. The lumped impedance model ‐ Linus............................................................... 10
2.1.4. The 1‐D model, the modified 1‐D model and the new model ‐ Linus ............... 10
2.1.5. The 3‐D FEM model ‐ Linus ................................................................................ 13
2.2. Pressure drop models ‐ Jesper...................................................................................... 14
2.2.1. 1‐D Pressure Drop Model ‐ Jesper ..................................................................... 15
3. Theory and measurement techniques .................................................................................... 18
3.1. Two‐port ‐ Linus............................................................................................................ 18
3.2. Microphone calibration ‐ Linus..................................................................................... 19
3.3. Pressure drop ‐ Jesper .................................................................................................. 20
3.4. Limitations ‐ Linus ......................................................................................................... 20
4. Measuring ‐ Jesper .................................................................................................................. 22 5. Results ‐ Linus .......................................................................................................................... 24
Transmission loss ......................................................................................................... 24
Pressure drop............................................................................................................... 26
6. Discussion – Linus & Jesper ..................................................................................................... 28 7. Conclusion ‐ Linus.................................................................................................................... 30 8. References............................................................................................................................... 31
Appendix A – Filter data Appendix B – Dimensions of the test rig Appendix C – Data sheet Bruel & Kjaer, Nexus model 2691‐A‐OS2 Signal Conditioner sn: 2301659 Appendix D – Microphone Preamplifier Type 2670 Appendix E – Data sheet Microphone ¼″ Pressure‐field Microphone — Type 4938, sn: 2323070, 2491541, 2239162, Brüel & Kjaer Appendix F – Calibrator Appendix G – Data sheet SwemaAir 300 Appendix H – Data sheet NAD C370, SR: H0ZC37006007 Appendix I – Gantt schedule