LAPPEENRANTA UNIVERSITY OF TECHNOLOGY LUT School of Energy Systems
LUT Mechanical Engineering
Vijaikrishnan Venkataramanan
DESIGN OF NOISE REDUCTION CAVITY STRUCTURES FOR ADDITIVE MANUFACTURING
Examiners: Professor Antti Salminen D. Sc. (Tech.) Heidi Piili
LUT Mechanical Engineering Vijaikrishnan Venkataramanan
Design of noise reduction cavity structures for additive manufacturing
Master’s thesis 2016
56 pages, 41 figures and 2 tables Examiners: Professor Antti Salminen
D. Sc. (Tech.) Heidi Piili
Keywords: Additive manufacturing, noise reduction, acoustic muffler.
Passive acoustic mufflers have a vital role in controlling the transmission of airborne noise passing into the human living environment. The geometry of the air path affects the attenuation of sound due to absorption and reflections of sound waves. Use of simple muffler geometry consumes a larger space for reduction of low frequency noises due to their longer wavelengths. A complex design of air cavity is useful to combine two or more sound absorption mechanisms to effectively reduce low and medium frequency noises in a minimal space. Manufacturing of complex cavity structures are simplified due to advent of additive manufacturing technologies.
A combined study on manufacturability of complex cavity structures and noise reduction performance due to geometric complexity remains relatively an unexplored field. This study aims to develop new designs for noise reduction cavity structures and numerically analyses the sound transmission loss through the cavities. The design rules used in this study incorporate principles of both noise reduction and additive manufacturing.
Study concludes design complexity of cavity a potential parameter that affects performance of passive noise reduction. Compact designs for additively manufacturable mufflers for controlling low and medium frequency noises are unveiled. Multiple expansion structure introduced in the study provides a peak transmission loss of 240 decibels and completely controls the sound less than 50 decibels for medium frequency range from 1000 Hz to 3250 Hz. Use of combination of Helmholtz cavity with multiple expansion structure provides a peak transmission loss of 173 decibels for a frequency of 620 Hz and completely controls the sound less than 50 decibels for a frequency range between 600 Hz to 1000 Hz. Increase of chamber width in multiple expansion structure widens the frequency bandwidth for noise reduction and decreases the peak transmission loss.
ACKNOWLEDGEMENTS
This study was carried out as a part of the Finnish Metals and Engineering Competence Cluster (FIMECC)’s program, MANU - Future digital manufacturing technologies and systems under the project, P6 next generation manufacturing.
I would like to express my gratitude to Professor Antti Salminen and D.Sc. Heidi Piili for their support and guidance to this thesis. I would like to thank especially my supervisor Heidi Piili for providing valuable thoughts and motivation from start to end of this study. I thank and acknowledge M.Sc. Mikko Hovilehto for his support and guidance for using FDM machine in LUT laser laboratory. I thank and acknowledge LUT IT- and Origo services for the software and facilities needful to carryout numerical simulations for this study. I would like to thank all the personnel of LUT Laser for providing a favourable laboratory environment for carrying out fabrication processes related to this study.
This thesis would have been impossible without the support and encouragement from my parents. I convey my thanks to all my friends and family members for their well-wishing thoughts about this work.
Vijaikrishnan Venkataramanan Lappeenranta, Finland 23.09.2016
TABLE OF CONTENT
ABSTRACT
ACKNOWLEDGEMENTS TABLE OF CONTENT
LIST OF SYMBOLS AND ABBREVIATIONS
1 INTRODUCTION ... 8
1.1 Airborne noise reduction ... 9
1.2 Scope and objectives of the study ... 10
1.3 Research questions ... 10
2 SOUND ABSORPTION ... 11
2.1 Effect of geometry ... 14
3 REDUCTION OF DUCT NOISE ... 19
3.1 Effect of multiple partition ... 20
4 ADDITIVE MANUFACTURING... 23
4.1 Design considerations in PBF process ... 24
4.1.1 Overhanging feature ... 24
4.1.2 Sharp edges ... 26
4.1.3 Gap between walls ... 27
4.1.4 Cross section area in building direction ... 27
4.1.5 Material accumulation ... 28
4.1.6 Accessibility ... 28
4.2 Manufacturing considerations in PBF process ... 29
4.2.1 Beam offset ... 29
4.2.2 Energy density ... 30
4.2.3 Contour scanning ... 31
5 AIM AND PURPOSE OF STUDY ... 32
6 DESIGN OF AIR CAVITY ... 33
7 STRUCTURE FOR AIR CAVITY ... 36
8 ACOUSTIC SIMULATION OF CAVITY ... 40
8.1 Mesh generation... 41
8.2 Code validation ... 43
5
8.3 Grid sensitivity test ... 44
9 RESULTS AND ANALYSIS ... 45
10 CONCLUSIONS ... 48
11 FURTHER STUDY ... 51
REFERENCES ... 52
LIST OF SYMBOLS AND ABBREVIATIONS
A Area of cross section in building direction [mm2] An Admittance [m/s.N]
bG Breadth of the gap between walls [mm]
c Velocity of sound in air [m/s]
[C]e Damping matrix for element
dA Differential area at duct entry and duct exit[mm2] Dm Diameter of capillary [mm]
Dp Diameter of perforation [mm]
{F}e Vector of acoustic force for element [N]
hG Height of gap between walls [mm]
i Complex number
[K]e Stiffness matrix for element
l Neck length of Helmholtz cavity [mm]
Lden Noise exposure indicator during day-evening-night [dB]
lG Length of gap between walls [mm]
lMA Section length in building direction [mm]
Lnight Noise exposure indicator during night [dB]
L1 Length of short path [mm]
L2 Length of long path [mm]
[M]e Mass matrix for element
n Normal direction
{N} Shape function of the element p Acoustic pressure [Pa]
Pin Acoustic power in the duct entry [W]
{pn} Acoustic pressure at nodes [Pa]
po Input acoustic pressure at the entry [Pa]
Pout Acoustic power in the duct exit [W]
q Double side extension length [mm]
r Single side extension length [mm]
R Fillet radius [mm]
7
S Neck area of Helmholtz cavity [mm2] Se Element surface area [mm2]
t Wall thickness [mm]
V Volume of Helmholtz cavity [mm3] Ve Element volume [mm3]
w Width of expansion chamber [mm]
Sound absorption coefficient f Frequency bandwidth [Hz]
L Difference between lengths of two paths [mm]
Angle of overhang [o]
’ Angle of overhang after deformation [o] Density of air [Kg/m3]
Diameter [mm]
Angular frequency [rad/s]
AM Additive Manufacturing
ABS Acrylonitrile Butadiene Styrene BEM Boundary Element Method CAD Computer Aided Design
DFAM Design for Additive Manufacturing FDM Fused Deposition Modelling FEM Finite Element Method
HC Helmholtz Cavity
MP Multiple Perforations PBF Powder Bed Fusion PH Precipitation Hardening
PLA Poly- Lactic Acid (Polylactide) STL Sound Transmission Loss
TL Transmission Loss
TMM Transfer Matrix Method
1 INTRODUCTION
Environmental noise level is an important factor affecting human health and comfort. Any unwanted sound is considered to be a disturbance for hearing and termed as noise. The ill effects of noise could be experienced by humans when hearing a varying amplitude sound above 35 dB (A). Disturbances due to noise level above 70 dB (A) prone to cause high stress, irritation, blood pressure, heart attack and other symptoms. (Miranda & Duarte 2008, p. 1.) The European environment agency reports that more than a 30% of the European population may be exposed to excessive noise levels causing living inconveniences. The chart in figure 1 shows the number of people exposed to different noise sources inside urban areas during day-evening-night times (Lden) and during night times (Lnight). (European Environment Agency 2014, p. 20.)
Figure 1.Number of people exposed to excess noise inside urban areas of Europe in 2007 (European Environment Agency 2014, p. 20).
9
Figure 1 shows that more than 100 million people are exposed to road side noise and more than 14 million people are exposed to railway noise. The traffic noise has been increasing every year by 0.2 to 0.3 dB (A) even in developed countries (Fuchs 2013, p. 1). The increasing number of transport vehicles, machineries and heavy industries urges the need for development of effective noise reduction treatments.
1.1 Airborne noise reduction
Rotary and repetitive machines, flow ducts with control valve, exhaust systems and ventilation systems are some of the common noise sources in living places that need effective treatment to reduce especially the low and medium frequency noises. The air path from the sound generating end is a crucial spot for sound reduction before it reaches the receiver. The airborne noise reduction can be achieved by sound energy absorption and multiple reflections along the air path. Figure 2 shows a vibrating object that generates sound energy which is carried in air field as sinusoidal oscillations (Barron 2003, p. 24). The vibration is transferred into air medium in a range of frequencies.
Figure 2.Viscous dissipation of airborne sound wave through hole (Barron 2003, p. 24).
As shown in figure 2, when the air flow approaches the solid surface, a viscous drag is created in the boundary. This results a viscous shear force between the air layers that restrict the oscillations created by sound. Finally the sound energy is damped and dissipated in form of heat. (Strek 2010, p. 664.) By allowing sound into a multiple partitioned air cavity made of dense wall material, the sound gets trapped and attenuated in air region inside the cavity
being reflected multiple times from walls due to large difference in sound impedance (Yu &
Cheng 2015, p. 01). A noise reduction structure is characterized by shape, size and geometric complexity of air cavity and the ratio of solid material to air. Manufacturing of complex cavity structures are made possible due to introduction of additive manufacturing (AM) technology. Designing and manufacturing of thin walled porous structures for increasing noise reduction performance remains relatively an unexplored field of study. Noise reduction efficiency of new air cavity designs and potential of additive manufacturing rise a need of combined research on both the fields. (Setaki et al. 2015, p. 02.)
1.2 Scope and objectives of the study
This study introduces complex air cavity structures as a noise control component which has same principle as that of passive acoustic silencers. The design of the cavity structure is useful for noise reduction treatment for exhaust ducts, ventilation ducts, automobile exhaust, valve contained pipelines and similar noise carrying ducts. The objectives of the study are:
To introduce a new design of complex air cavity for noise reduction;
To utilize finite element method (FEM) to solve acoustic pressure through three dimensional cavity and to calculate sound transmission loss (STL);
To analyse the effect of design parameters on noise reduction performance across the cavity and
To incorporate AM procedure for both design and manufacturing of complex structure for the cavity.
1.3 Research questions
This study is intended to give answers to following research questions:
1. How to reduce airborne noise using structural complexity?
2. How to utilize AM technology to manufacture complex cavity structure?
11
2 SOUND ABSORPTION
The propagation of sound energy through air can be characterised by consecutive compression and expansion motion of air starting from the source point till its complete dissipation into heat energy. The compression and expansion motion is affected when there is a velocity difference between two adjacent air molecules. The relative velocity between the molecules can be varied inside the medium itself due to intervention of foreign molecules, for example humidity or other chemical agents (Bohn 1988, p. 06). In another way, the velocity difference is also created by introducing solid surface obstacles in the air path. In such cases, the velocity of air molecules adjacent to solid boundary is zero and the next layer of air molecules travel with a finite velocity. (Rienstra & Hirschberg 2016, p. 79.) As shown in figure 3, the variation in velocity continues layer by layer till it reaches the free flow region or maximum velocity region (Godbold 2008, p. 15).
Figure 3.Viscous boundary during solid fluid interaction (Godbold 2008, p. 15).
Figure 3 shows the variation in velocity vector from solid boundary to free flow region. The frictional force created between two layers acts as disturbance to compression and expansion motion of air reducing its kinetic energy and dissipating it into heat. By increasing the solid- fluid contact surface area or by increasing the amplitude of oscillations near by the surface, the effect of viscous dissipation of sound is further increased (Godbold 2008, p. 15; Strek 2010, p. 664). The contact surface can be increased when sound propagates through perforated enclosures as shown in figure 4. The effect of sound absorption can be increased
by introducing structures with continuous pores in the sound propagation path (Geslain et al. 2012, p. 208).
Figure 4.Cross section view of porous material (Arenas & Crocker 2010, p. 12).
It can be seen from figure 4 that some pores are not continuous and some of them are continuous and can allow propagation of sound. When air carries a particular wavelength of sound through hole, the maximum velocity occurs a quarter of wavelength from the wall surface. The depth of the sound carrying hole has to be more than a quarter of the wavelength of incident sound to achieve effective sound absorption. (Hannink 2007, p. 05.) In order to absorb a wave with frequency of 1 kHz, the depth of the pore has to be more than quarter its wavelength, i.e. > 8.5 mm. When absorbing low frequency wave for example 50 Hz, the quarter wavelength is 1700 mm. In such cases, using longer pored absorbers has practical difficulties due to its larger size. Another mechanism of sound absorption could be achieved by increasing the oscillation of sound adjacent to solid surfaces. This principle is followed in Helmholtz Cavity (HC) absorbers as shown in figure 5 (Klaus et al. 2012, p. 489).
13
Figure 5.Basic Helmholtz cavity structure (Godbold 2008, p. 18).
Figure 5 shows a simple Helmholtz cavity of volume V, neck lengthl and neck areaS. When an incident sound wave approaches the opening, the air in neck region oscillates to and fro and the air inside the cavity compresses to some extent. The amplitude of oscillation is increased to maximum when the incident waves are with same frequency that of oscillating air plug which is also called as resonance frequency. Higher the amplitude of oscillation, greater is the velocity difference between air layers that will result in maximum viscous dissipation of sound (Biswas & Agrawal 2013, p. 1681).
Passive destructive interference is another concept of cancelling the sound wave oscillations.
This involves in interference of two sound waves with a phase difference of 180o(see the figure 6). When sound waves of same wavelength travel in two different path meet each other, there is a possibility of complete destructive interference. This depends on length of travel in two different paths and wavelength of the incident wave. (Setaki et al. 2014, p. 189.)
Figure 6.Principle of passive destructive interference absorber (Setaki et al. 2014, p. 189).
Figure 6 shows the relationship between length of path 1,L1 and path 2,L2 to achieve 180o phase difference at the point of interference. It is possible when the difference in length between two paths, Lis equal to half the wavelength of the incident sound wave. Advantage of this construction is that a same source of sound is used for two different paths and no additional sound source is necessary. (Setaki et al. 2014, p. 189.)
2.1 Effect of geometry
A cylindrical cavity with orifice is analysed by Godbold (2008) to study the effect of geometry on sound absorption performance. The resonance frequency at which there is a peak value of absorption and the frequency bandwidth at half the peak of absorption coefficient, , are the two figures of merit considered. The orifice diameter and orifice length are varied from a mean value of 15 mm and its effect on absorption parameters are studied as shown in figure 7. (Godbold 2008, p. 75.)
15
Figure 7.Effect of varying orifice dimensions (Godbold 2008, p. 75).
Figure 7 shows that increasing the length of the orifice reduce both peak frequency and the frequency bandwidth at half the peak of absorption coefficient, in a higher rate up to 5 mm and in lower rate beyond 15 mm length. It is understood that higher the orifice diameter, the peak absorption takes place in higher frequencies with a higher bandwidth at half the peak absorption. Changes in shape and internal surface area of the cavity by including internal structures provide further possibility to increase sound absorption. The effect on sound absorption with fin arrays inside the cavity is shown in figure 8. (Godbold et al. 2007, p. 04.)
Figure 8.Absorption of cavity with fin array (Godbold et al. 2007, p. 04).
Figure 8 shows that the peak absorption increases by decreasing the fin spacing and in the absence of fins, the absorption is the least. The viscous dissipation of sound increases with smaller fin spacing that lead to maximum absorption due to larger contact with wall area (Strek 2010, p. 664). When compared to the empty cavity, there is a shift of peak frequency from 300 Hz to 287 Hz and the bandwidth at half the peak of absorption coefficient, (at 0.43) is increased to 78.13 Hz due to the presence of internal fins with a spacing of 2 mm. When the fin spacing is further decreased, the peak frequency shifts towards a higher frequency value (Godbold et al. 2007, p. 04). A similar effect on sound absorption is inferred in case of micro capillary films. Absorption of sound in micro capillary path is studied by using perforated micro capillary films as shown in figure 9. (Xu et al. 2015, p. 154–156.)
Figure 9.Micro capillary film with perforations (Xu et al. 2015, p. 154).
Figure 9 shows the arrangement of micro capillary films orthogonal to each other and the perforations are formed normal to film. Where, Dp is the diameter of perforation and Dm is the diameter of capillary. The first phase of absorption takes place through the perforations and the second phase of absorption occurs through micro capillaries. An eight layer of perforated micro capillary films with 28 capillaries in each layer is compared with a similar structure having one capillary in each layer. The perforation diameter of 2 mm with a cavity depth of 50 mm is maintained in both the cases. The results of sound absorption is shown in figure 10.
17
Figure 10.Effect of number of capillary paths on sound absorption (Xu et al. 2015, p. 155).
Figure 10 shows that absorption coefficient is increased for majority of frequencies in case of using 28 micro capillary paths except for 200 Hz. There is a four times increment in absorption for the frequency equal to 2000 Hz. The peak absorption is the highest when the number of capillaries is more and similar is the effect in the study made by Godbold et al.
(2007), when the number of fins are more with lesser spacing to each other. (Godbold et al.
2007, p. 04; Xu et al. 2015, p. 154.) The effect on absorption coefficient by mounting Helmholtz cavity on perforated chamber is studied by Gai et al. (2016) shown in figure 11.
Figure 11.Effect on Helmholtz cavity mounted perforated chamber (Gai et al. 2016, p. 262).
Figure 11 shows the absorption coefficient due to separate and combined effect of Helmholtz cavity and multiple perforated (MP) chamber. Helmholtz cavity has higher absorption coefficient for narrow band low frequencies. The perforated chamber has the absorption for a wider band of frequency. The total bandwidth of absorption widens further by combining both Helmholtz cavity and perforated chamber. (Gai et al. 2016, p. 262; Park 2013, p. 4900.) The arrangement that contain a perforated chamber inside Helmholtz cavity is studied by Godbold (2008). A perforated plate having a hole diameter of 0.5 mm is placed inside the cavity and its absorption coefficient is compared with empty Helmholtz cavity as shown in figure 12. (Godbold 2008, p. 75.)
Figure 12.Effect on perforated chamber inside Helmholtz cavity (Godbold 2008, p. 75).
Figure 12 shows that the absorption is on low frequency region for a narrow bandwidth which is due to Helmholtz cavity (Gai et al. 2016, p. 262; Godbold 2008, p. 75; Park 2013, p. 4900). The presence of perforated plate shifts the peak frequency from 300 Hz to 291 Hz and increases the bandwidth at half the peak of absorption coefficient, ( ). This is due to increased wall contact area and viscous dissipation of sound. (Godbold 2007, p. 05; Strek 2010, p. 664; Xu et al. 2015, p. 156.)
19
3 REDUCTION OF DUCT NOISE
Noise transmission through ducts is critical in exhaust system of combustion engines, power production industries, ventilation and air-conditioning, etc. Noise reduction through ducts can be achieved using acoustic silencers in the sound transmission path. Silencers are classified into active and passive type. The active silencers produce sound field which is a mirror image of disturbing sound field. The two opposite fields cancel each other during interference and reduces the noise level. Passive acoustic silencers are classified into three types namely dissipative, reactive and hybrid based on their noise reduction principle.
Dissipative silencers follow principle of absorbing the sound and dissipating it into heat.
Reactive silencers attenuate the sound by reflections caused due to impedance difference in the medium. Hybrid silencers follow both dissipative and reactive principles for effective noise control. (Yu & Cheng 2015, p 3.) A general design methodology for acoustic silencers is given by Yu and Cheng (2015) which is shown in figure 13.
Figure 13.General silencer design methodology (Yu & Cheng 2015, p 3).
Figure 13 illustrates a flow chart for silencer design containing three different stages. The preparation stage involves identification of incoming noise properties such as frequency, amplitude, angle of incidence and idea about target noise level in the receiver end. In the
second stage, an initial design criteria, for example, a sub-chamber based silencer design (Yu & Cheng 2015, p 1) or a multi cavity muffler (Ouédraogo et al. 2016, p 21) or any other criteria is decided and the range for geometric parameters are set. An acoustic analysis is carried by using mathematical methods for example, finite element method (FEM) (Qian et al. 2015, p. 87) or transfer matrix method (TMM) (Vigran 2012, p. 454), etc., to study the effect of geometry on noise reduction performance. Optimum values of geometric parameters are identified and further analysis is carried according to working condition. In the realization stage, prototypes are created and noise reduction performance is tested experimentally. (Yu & Cheng 2015, p 4.)
3.1 Effect of multiple partition
A cascade silencer made by connecting sub chamber configurations is studied by Yu and Cheng (2015). Three types of expansion chambers as shown in figure 14 are designed and analysed individually for noise reduction performance.
Figure 14.Three types (a), (b), (c) and a 3D view of type (b) (Yu and Cheng, 2015, p 1).
As shown in figure 14, the chamber width w, single side extension lengthr and double side extension length q are considered as geometric variables. When increasing chamber width from 0.05 m to 0.1 m, it is observed that the peak transmission loss occurs in higher frequency region. By increasing the length r from 0.03 m to 0.07 m, the peak transmission loss shifts to a low frequency region due to increase in characteristic height of chamber and there is also an increase in bandwidth at transmission loss value of 20 dB. In case of increasing length of double sided extensions,q, the bandwidth at 20 dB transmission loss is reduced and there is a shift of peak transmission loss towards lower frequency region. (Yu
& Cheng 2015, p 2.) Effect of cascading all the three sub chambers added with one more chamber of type (a) on sound transmission loss is illustrated in the figure 15 (Yu, Tong, Pan
& Cheng 2015, p. 66).
21
Figure 15. Performance of individual and cascaded chambers (Yu, Tong, Pan & Cheng 2015, p. 66).
Figure 15 shows the transmission loss for four individual sub-chambers and a cascaded chamber in the order 1, 2, 3, 4: Chamber 1 is of type (c) with q = 0.07 m, chamber 2 is of type (b) with r = 0.06 m, chamber 3 and 4 are of type (a) with w = 0.08 m and 0.125 m respectively. It can be noted that the performance of cascaded chamber is remarkably better than that of individual chambers. By considering a target transmission loss of 20 dB, an efficient blocking of sound from 280 Hz to 1380 Hz is achieved and by increasing the target level to 40 dB, a complete blocking is achieved from 500 Hz to 1350 Hz as a result of cascading the sub-chambers. (Yu, Tong, Pan & Cheng 2015, p. 66.)
Multiple partitioning of sound passage is also applicable for the sound flow through orifice.
Noise transmission through a duct with an orifice is studied by Qian et al. (2015). The orifice with single flow passage without any obstacle and orifice with multiple flow passage guided by a thick perforated plate are compared. The total area of cross section of multiple flow passage is maintained equal to that of single flow passage. The effect of splitting the flow passage on sound transmission loss is examined as shown in figure 16. (Qian et al. 2015, p.
90.)
Figure 16.Sound transmission via orifice and perforated plate (Qian et al. 2015, p. 90).
It can be noted from the figure 16, that the frequency bandwidth, f in which there is a continuous transmission loss (TL), is increased from 2750 Hz to 3100 Hz by replacing the single flow orifice passage with thick perforated plates guiding multiple flow passages. It can be understood from the above two examples, that by creating partitions along or across the sound flow passage, a positive effect on noise reduction could be achieved. There is relatively a wider scope for design development of duct acoustic silencers and mufflers to increase its noise reducing performance and compactness. This may give a simpler solution for noise reduction compared to modifying the properties of duct or fluid. However, there are only limited number of literature dealing with studies of the effect of geometrical variations of structures on noise reduction performance. (Qian et al. 2015, p. 87.) As the manufacturing freedom of complex designs has been increased by AM technologies, a better solution for noise reduction could be achieved by utilizing the manufacturing capability of AM.
23
4 ADDITIVE MANUFACTURING
Additive manufacturing is a manufacturing technology that follows principle of adding material mostly in a layer by layer fashion to build a component. Many variants of equipment had been developed in additive manufacturing family in the last three decades. The capabilities and limitations of equipment variants differ from each other in terms of material, accuracy, geometric complexity, mechanical properties and economic consideration.
Powder bed fusion (PBF) (ASTM F2792-12a 2013, p. 2) is one of the AM processes in which almost fully dense material can be built up layer after layer to consolidate parts of definite geometry as shown in figure 17 (Scotti et al. 2016, p. 476).
Figure 17.Powder bed fusion process (Scotti et al. 2016, p. 476).
As shown in figure 17, a 3D model developed by computer aided design (CAD) is sliced into 2D layers and used as input geometry for fabrication. A numerically controlled laser beam guided by scanner optics is used to melt each layer. The powder bed platform lowers to a distance equal to exact layer thickness and fresh powder is evenly spread over the platform after melting every layer. A main advantage of PBF process lies in economic manufacturing of parts with higher level of complexity. This process is relatively slow compared to other AM processes. Fused deposition modelling (FDM) (ASTM F2792-12a 2013, p.2) is one more commonly used AM process especially for acrylonitrile butadiene styrene (ABS) and poly-lactic acid (PLA) polymers in which the material is added through
nozzle layer by layer with a constant pressure as shown in figure 18 (Novakova- Marcincinova 2012, p. 36).
Figure 18.Fused deposition modelling (Novakova-Marcincinova 2012, p. 36).
As shown in figure 18, the material filaments are fused by heated nozzles and deposited layer by layer to build the main part and support structure. The accuracy of part depends on the radius of the nozzle opening. The layer thickness typically ranges from 0.18 mm to 0.35 mm.
(Kannan & Senthilkumaran 2014, p. 1048) Higher layer thickness imposes limitations while building the parts with higher geometrical complexity and thinner structures. Thus PBF process is applicable for manufacturing a wider range of complex designs which are useful for building complex cavity structures for noise reduction. In case of building thin wall structures in PBF process, the quality of part depends on the form, dimensional accuracy and porosity. (Abele et al. 2015b, p. 119; Gu & Shen 2008, p. 1884) It is essential to consider certain procedures in design and manufacturing for PBF process in order to avoid manufacturing failure and to ensure part quality.
4.1 Design considerations in PBF process
Design for additive manufacturing (DFAM) follows certain rules during part design creation to improve the ease of manufacturing through PBF process. Serious defects like thermal distortion, breakage and warping can be prevented by careful designing. Improvement in surface quality, ease of support removal and extraction of unexposed powder from the cavities could also be ensured through design.
25
4.1.1 Overhanging feature
Overhangs are the structures built directly over the stock of powder that may be inclined to certain angle from the base plate as shown in figure 19 (Wang et al. 2013, p. 1740).
Figure 19.Schematic diagram of overhanging surface (Wang et al. 2013, p. 1740).
Figure 19 shows the overhanging surface in the region of fabricated layers adjacent to powder region with an angle of inclination from the building direction. The heat from laser beam is conducted to supporting powder zone in a very slow rate, equal to 1/100 times to that of conduction in solid supporting zone. (Kruth et al. 2007, p. 03; Wang et al. 2013, p.
1738.) Due to poor conduction, the heating caused by absorption of laser irradiation is much higher in the regions supported by powder. This causes the formation of larger molten pool that intrudes further into the powder region due to gravitational and capillary forces. As a result, there will be excessive powder adhesion from bulk powder region leading to inaccuracies in part dimension especially at the lower side of overhang. When the thermal stress is greater than the strength of the material, warping of layer takes place as shown in figure 20. (Liu, Yang & Wang 2016, p. 07; Wang et al. 2013, p. 1739.)
a) b)
Figure 20.Warping single layer (a) and cumulative layers (b) (Wang et al. 2013, p. 1739).
Figure 20 shows the principle of warping due to thermal stress. Due to constriction towards solid region, the extended surface warps in upward direction. The designed angle of overhang increases to ’ due to warping of cumulative layers. When the warped height exceeds the layer thickness, the recoating of new layer is affected. In the worst cases, the part may hit the recoater and got dragged by its motion collapsing the part structure. The severity of powder adhesion and layer warping depends on the angle of overhang and length of extension. It has been tested for the material 316L stainless steel powder, that when >
40o, the design lead to stable fabrication and when < 35o the manufacturability is poor due to high risk of powder adhesion and layer warping. (Wang et al. 2013, p. 1736.) The overhang structure remains stable even at = 0o if the length of extension is less than 2 mm (Adam & Zimmer 2014, p. 27).
4.1.2 Sharp edges
Sharp edges are not possible to build accurately and there will be a formation of curvature depending on laser spot diameter (Adam & Zimmer 2014, p. 26; Kruth et al. 2007, p. 03) (See the figure 21).
Figure 21.Design for outer edge (a) inner edge (b) (Adam & Zimmer 2014, p. 26).
Figure 21a shows outer edges built in both vertical and horizontal direction. Zrepresents the building direction. It is favourable to build blunt surfaces with certain dimension in the place of outer sharp edges. As shown in figure 21b, the downward facing inner edges would be left sharp in order to avoid building support structures. In case of upward facing inner edges,
27
a minimum radius is recommended in order to facilitate removal of unexposed powder from the edge. (Adam & Zimmer, 2014, p. 26.)
4.1.3 Gap between walls
The dimensions of gap between two adjacent walls are chosen in order to allow easy removal of unexposed powder (Adam & Zimmer 2014, p. 27; Thomas, 2009, p. 164) (See the figure 22).
Figure 22.Dimensions of gap between walls (Adam & Zimmer 2014, p. 27).
Figure 22 shows length, lG, height, hG and breadth of the gap, bG. From the test conducted by Adam and Zimmer (2014), it is recommended to maintain hG>0.2mm for lengthlG 50 mm. The breadth bG can be chosen freely. It is safe to maintain 0.3 mm distance between two different features (Thomas 2009, p. 164). The minimum thickness of wall for stainless steel L316 is recommended to be at least 0.4 mm to 0.6 mm for a defect free end component.
(Adam & Zimmer 2015, p. 669; Thomas 2009, p. 166.)
4.1.4 Cross section area
A decreasing area of cross section of part from top to bottom is not favourable to achieve good manufacturability (Adam & Zimmer 2014, p. 27) (See the figure 23).
Figure 23.Change in cross section area (Adam & Zimmer 2014, p. 27).
Figure 23 shows that the area of cross section of upper segmentA3 is to be less than the total cross section area of lower segmentsA1 andA2 for favourable manufacturing condition. The heat conduction rate increases due to increase in cross section area that reduces the accumulation of thermal stress on the part. (Adam & Zimmer 2014, p. 27.)
4.1.5 Material accumulation
Stability of material accumulation is tested by Adam and Zimmer (2014) by aggregating two elements each other (See the figure 24).
Figure 24.Material accumulation (Adam & Zimmer 2014, p. 27).
Figure 24 shows that material aggregated with smaller cross section length, lMA is more favourable. Larger cross sections result in defects due to thermal stress. It is recommended to maintainlMA 20mm for good manufacturability. (Adam & Zimmer 2014, p. 27.) 4.1.6 Accessibility
While designing complex structures, there should be enough space to access internal supports and powder inside a built up cavity (Kranz, Herzog & Emmelmann 2015, p. 8, 12) (See the figure 25).
Figure 25.Accessibility of powder and support structure (Kranz et al. 2015, p. 8, 12).
29
Figure 25 shows that it is recommended to provide at least one open side for the support removal and to guarantee ease of final machining. In case of complex cavities, more than one vent opening is favourable for the ease of powder removal. (Kranz et al. 2015, p. 8.) 4.2 Manufacturing considerations in PBF process
Manufacturing practices may affect the quality of thin wall structures in terms of dimensional accuracy, porosity and surface roughness. Procedures such as beam offsetting and contour scanning would improve the part dimensional accuracy. Maintaining an optimum value of laser energy density could minimize the porosity and increase the part density.
4.2.1 Beam offset
The dimensional accuracy of thin wall structures is affected by varying the beam displacement from the designed profile (See the figure 26a) (Abele et al. 2015a, p. 119). The discrepancies in dimensions of manufactured profile can be prevented by introducing beam offset from designed profile (See the figure 26b) (Abele et al. 2015b, p. 117).
a) b)
Figure 26.Laser beam displacement (a) (Abele et al. 2015a, p. 119) and deviation in wall thickness (b) (Abele et al. 2015b, p. 117).
In figure 26a, the path of laser beam spot is displaced towards inward direction from original profile such that the outer edge of laser beam coincides with the profile. The energy from
the laser is almost directed into the designed profile thus avoiding the melting of powder outside the profile. The value of beam displacement has been experimentally determined by Abele et al. (2015b). Figure 26b shows the measurements of wall thickness from experimental results obtained before and after beam compensation. An average deviation of 171 µm from the designed profile has been observed for the material, precipitation hardening (PH) 17-4 stainless steel in the initial measurements before making beam offset. The value of beam offset has been considered approximately half of the average deviation from initial measurement, i.e. 86 µm. It can be observed from the measurement taken after beam compensation that the dimensional discrepancies are effectively controlled. (Abele et al.
2015b, p. 117.)
4.2.2 Energy density
Laser energy density is the amount of energy incident on unit area that can be adjusted by varying the parameters such as laser scanning speed, hatch distance and laser power. The energy density affects porosity of part in an inverse relationship as illustrated (Abele et al.
2015b, p. 118) (See the figure 27).
Figure 27.Effect of energy density on porosity of part (Abele et al. 2015b, p. 118).
31
It can be noted from figure 27 that the minimum percentage of porosity is 0.99% that is obtained for energy density of 1.9 J/mm2. By decreasing the energy density up to 0.57 J/mm2, the porosity percentage is raised to 17.35%. From the regression model developed by Abele et al. (2015) for the material PH 17-4 stainless steel, the major factor contributing for the effect of energy density on porosity is found to be the hatch distance (77.63%) and secondly the scanning speed (14.82%). The laser power has a lesser contribution to the effect on porosity which is equal to 3.04%. (Abele et al. 2015b, p. 118.)
4.2.4 Contour scanning
Contour scanning is a method of scanning the profiles (edges) of the part where the laser beam follows the contour of the profile. It helps in proper fabrication of outer and inner profiles by avoiding the geometric defects in edges. Effect of contour scanning is studied by Su et al. (2012) using the material 316L stainless steel for an optical component (See the figure 28). (Su et al. 2012, p. 1236.)
Figure 28.Contour scanning and geometric defect (Su et al. 2012, p. 1236).
Figure 28 shows a schematic of inter-layer stagger scanning with and without contour scanning. It can be noted that there is a formation of discontinuous outer profile without contour scanning. It is due to absence of molten material between the track A and track B at the outer edge of the part. The interior hatching may increase the relative density of the part but a re-melting of profiles using contour scanning is necessary to achieve the accurate shape. (Su et al. 2012, p. 1236.)
5 AIM AND PURPOSE OF STUDY
The aim of the study is to develop new design for passive acoustic mufflers that has higher value of sound transmission loss for low and medium frequency noises. The effect of geometrical changes of air cavity on its noise reduction performance is studied. The air cavities are modelled and analysed using finite element method to calculate sound transmission losses. The design of cavities and their structures are developed by utilizing the design freedom of additive manufacturing.
33
6 DESIGN OF AIR CAVITY
Two configurations of multiple partitioned cavities are proposed for the study. The key noise reduction elements are a multiple expansion structure, a single expansion chamber and Helmholtz chamber which are shown in figure 29 and figure 30. Figure 29 illustrates the design of the first configuration in both combined view and split view.
Figure 29. Combined view and split view of air cavity configuration-1.
The split view in the figure 29 shows the sequence of sound transmission from duct entry (1) to duct exit (6). The input planar sound wave from the duct entry is carried by inlet (2) which is augmented with multiple expansion structure (3) where the width of each expansion chamber is 1 mm. The resultant sound from the inlet (2) passes inside a single expansion chamber (4) followed by neck (5) and finally to duct exit (6). Another configuration is proposed with integration of a Helmholtz chamber. Figure 30 illustrates the design of the second configuration in both combined view and split view (See the figure 30).
Figure 30. Combined view and split view of air cavity configuration-2.
35
Figure 30 shows the sequence of sound transmission of the proposed configuration-2. The sound wave from the duct entry (1) passes through the inlet (2) and simultaneously spread through multiple expanding structure (3). The resultant sound wave enters the Helmholtz chamber (5) through the slit openings (4) and enters back into (3) through the openings (6) and finally reaches the outlet (7) and escapes through the duct exit (8).
7 STRUCTURE FOR AIR CAVITY
The structure for both the air cavity configurations is modelled using the tool, SolidWorks (Dassault Systems SolidWorks corp., 2015). The dimensions of the cavity features are maintained the same as described in chapter 6. The structures are designed considering the ease of additive manufacturing processes such as FDM and PBF. Figure 31 shows the structure for the cavity configuration-2.
Figure 31.Structure for cavity configuration-2.
37
Figure 31 shows the main sections of the structure for cavity configuration-2 in a trimetric view. The cavity is oriented in such a way the overhangs are inclined 45o from horizontal plane thus requirement of support structure is avoided for the overhangs. The minimum spacing between the adjacent walls and the minimum wall thickness (t) are kept 1 mm. The minimum fillet radius (R) is kept 0.4 mm to avoid sharp edges. Similarly, the structures for configuration-1 including and excluding the single expansion chamber are modelled as shown in figure 32a and figure 32b.
Figure 32.Structure for cavity configuration -1. (a) With single expansion chamber.
(b) Without single expansion chamber.
Figure 32 shows the main section of the cavity structures for cavity configuration-1 in a trimetric view. As shown in figure 32b, all the overhangs are inclined to 45o and support structures are avoided. For both configuration-1 and configuration-2, the side walls of
bottom portion have diverging cross section such that the area increases towards the base.
The wall helps the heat conduction from all parts of the structure. The increasing area towards the base increases the rate of heat conduction that ensures free flow of heat and avoids distortions. The structure for configuration-2 is built by fused deposition modelling as shown in figure 33a using desktop 3D printer in LUT laser laboratory. The completely built-up prototype without surface smoothening is shown in figure 33b.
(a)
(b)
Figure 33.Structure for configuration-2. (a) 12% completion. (b) 100% completion.
39
Figure 33a shows the setup for 3D printing and the status of job in 12% of its completion.
Filament made of polylactide (PLA), biodegradable plastic material is used for deposition.
The temperature inside nozzle is maintained to be 180oC to fuse the material to the printable state. The thickness of each layer of deposition is maintained to be 0.02 mm. As shown in figure 33a, the thicker walls are built porous with internal hexagonal structure by setting 30% infill percentage in the program. The design of component 100% supported the manufacturability by fused deposition modelling. Use of selective laser sintering and direct metal laser sintering is expected to have no risks for manufacturing of the designs given in this study. The plastic printed structures can be readily used for domestic purposes as window ventilators. Further analysis is required for industrial and automotive applications based on the working conditions.
8 ACOUSTIC SIMULATION OF CAVITY
The acoustic pressure across the cavity is calculated by solving a modified Helmholtz equation using finite element method with help of software, COMSOL Multiphysics. The boundary conditions are considered for sound wave entry region, walls and the sound wave exit region.
2
. p 2 p 0
c (1)
. 2 o
p i i
n p p
c c (2)
. 0
p n (3)
p . i
n p
c (4)
2
2
o in
P p dA
c (5)
. ( )
out 2
p conj p
P dA
c (6)
10.log in
out
TL P
P (7)
Where,p is acoustic pressure, is angular frequency, is density of air (1.225 kg/m3),c is velocity of sound in air (343.2 m/s), po is input acoustic pressure,n is the normal direction,
41
iis the complex number, TL is transmission loss of sound, Pin is acoustic power in the duct entry,Pout is acoustic power in the duct exit and dA is the differential area at entry and exit.
Equations 1 to 7 are referred from COMSOL software documentation (COMSOL Inc., 2015). Equation 1 is the governing equation for solving pressure field. The input plane wave boundary applied at duct entry is a combination of both incident and reflecting pressure waves which is given by equation 2. Sound hard wall boundary, given in equation 3 is applied to the interior and exterior walls implying no transmission of sound through walls.
An outgoing plane wave boundary, given in equation 4 is applied to duct exit. From pressure values obtained from the solution, the acoustic power at the entry and exit of the cavity are calculated using equations 5 and equation 6. Finally the sound transmission loss across the cavity is calculated by substituting the acoustic power values in equation 7.
8.1 Mesh generation
The three dimensional domain of the air cavity is divided into number of finite tetrahedron elements that contain four nodes in its corners as shown in figure 34.
Figure 34. Meshed domains using 4-node tetrahedron element.
Figure 34 shows a non-uniform mesh pattern using 4-node tetrahedron elements. The size of the element is maintained relative to the dimension of the feature and the mesh density is maintained high in the places of sudden change in area. An algebraic equation for an individual tetrahedron element is formulated by Qian et al. (2016) in terms of nodal acoustic pressure induced by external acoustic force and the resistance created due to factors such as mass, stiffness and damping property of air medium. (Qian et al. 2016, p. 87.)
e e 2 e e
K i C M pn F (8)
[ ]e e T e
K V N N dV (9)
[ ]e n e T e
C A S N N dS (10)
2
[ ]e 1 e T e
c V
M N N dV (11)
e
e e
S o
F p N dS (12)
The equation for a finite element is shown in equation 8. The matrices and vectors are given in terms of shape function of the element in the equations 9-12. Where, [K]e is the stiffness matrix,i is the complex number, is the angular frequency, [C]e is the damping matrix, [M]e is the mass matrix, {pn} is the acoustic pressure at nodes and {F}e is vector of acoustic force, Ve is the element volume, {N} is the shape function of the element, is the density of air,An
is the admittance,Se is the element surface area,c is the velocity of sound in air andpois the applied acoustic pressure in sound entry region. As the force vector is only applicable in entry and exit regions of sound, it has a null value in all internal and wall region. The single algebraic equation of all the elements in meshed domain is formulated by assembling individual element equations and later it is solved using direct sparse matrix solution method to find the unknown nodal pressure values {pn} of all the elements. (Qian et al. 2016, p. 87.)
43
8.2 Code validation
In order to validate the COMSOL code used for finite element analysis of proposed cavity, a simple expansion chamber is solved and the result is compared with previous experimental result referred from Tao and Seybert (2003). The diameter, in different section of the chamber (in inches) and transmission loss (TL) calculated using both experimental method and boundary element method are given in figure 35 (Tao & Seybert 2003, p. 4). The results from COMSOL code for solving the same problem is given in figure 36.
Figure 35.Transmission loss for expansion chamber (Tao & Seybert 2003, p. 4).
Figure 36.Transmission loss for simple expansion chamber calculated using COMSOL.
From figure 35 and figure 36, the transmission loss calculated with the procedure used in this study can be compared to results from experimental method and boundary element method (BEM), obtained by Tao and Seybert (2003). It can be noted that the calculated transmission loss for frequency range 0 Hz to 3000 Hz are in good agreement with experimental results with a deviation roughly less than 5%. Thus the solution procedure used in this study is validated and used for solving the proposed air cavity configurations.
8.3 Grid sensitivity test
The effect of element size on the solution is examined by varying the maximum and minimum element sizes in the meshed domain. The deviation in transmission loss value at the peak frequency due to change in element size is observed. Table 1 shows the result of grid sensitivity test for the proposed configuration 1.
Table 1. Grid sensitivity test Sl. No Maximum size
(mm)
Minimum size (mm)
Transmission loss (dB)
1 42.20 8.88 172.97
2 33.30 6.20 172.20
3 22.20 4 171.61
4 17.80 2.22 170.96
5 12.20 2 170.84
6 7.7 2 170.77
Table 1 shows that the transmission loss value decreases with decrease in element size. As the effect of element size is less significant, a shorter computational time becomes next interest. So that a medium grid size of 22.2 mm (maximum) and 4 mm (minimum) is chosen for the study.
45
9 RESULTS AND ANALYSIS
The dimensions of key elements such as single expansion chamber, Helmholtz chamber and multiple expansion structure are varied and the performance of noise reduction is analysed.
Three design cases as described in table 2 is analysed.
Table 2. Three design cases based on width of cavity
Sl. No Description Configuration
Case1 Multiple expansion chamber for 100% width of cavity 1 Case 2 Single expansion chamber for 50% width of cavity and
multiple expansion chamber for 50% width of cavity.
1
Case 3 Helmholtz chamber concentric with multiple expansion chamber that cover 100% width of cavity.
2
The sound transmission loss resulted in all the three cases is compared in figure 37.
Figure 37.Effect of design changes on sound transmission loss.
Figure 37 shows the sound transmission loss obtained in three different design cases from which the effect of key elements of proposed configurations on noise reduction can be observed. The STL is the maximum for the design case 1 with the peak value of 240 dB and
it is more than 50 dB for the frequency range from 1400 Hz to 3250 Hz. The peak value is reduced in design case-2 due to inclusion of single expansion chamber for half the width of the cavity replacing a portion of multiple expansion structure. It can be observed by comparing case-1 and case-2 that the use of multiple expansion structure gives a larger performance for noise reduction for higher frequencies. This is due to increased area of contact and high viscous dissipation of medium and high frequency sound waves. The performance due to use of single expansion chamber in case-1 is better for low frequency noises from 500 Hz to 1420 Hz because of its larger size. In order to get further higher STL for low frequency sound input, a Helmholtz chamber is included in design case-3. By combining a multiple expansion structure with Helmholtz chamber, a maximum STL of 173 dB is achieved at the frequency of 620 Hz and more than 50 dB transmission loss is obtained for the frequency range from 600 Hz to 1000 Hz. The effect of chamber width in multiple expansion structure on STL is studied for case 3 in presence of Helmholtz chamber. The results are compared in figure 38.
Figure 38.Effect of chamber width in multiple expansion structure for case-3.
Figure 38 shows that the maximum sound transmission loss occurs when the chamber width is kept 1 mm. The performance is reduced when chamber width is increased to 10 mm and to 21 mm. This is due to larger number of expansions at lower chamber width that cause increased area of contact and higher viscous dissipation of sound. Due to the combined effect of Helmholtz chamber and multiple expansion structure, the STL increases for the majority
47
of lower frequencies between 10 Hz to 600 Hz with decrease in chamber width. In order to observe the effect of chamber width in multiple expansion structure in absence of Helmholtz chamber, the case-1 is evaluated as shown in figure 39.
Figure 39.Effect of chamber width in multiple expansion structure for case-1.
Figure 39 shows that the peak value of STL is decreased with increase in chamber width in multiple expansion structure from 1 mm to 20 mm. In the absence of Helmholtz chamber, the STL for low frequency noises increases with increase in chamber width. But the noise reduction performance in the low frequencies is much lesser than that compared to case-3.
It can be inferred that the use of multiple expansion structure alone gives highest noise reduction performance for medium frequency noises ranging from 1000 Hz to 3250 Hz. The use of Helmholtz chamber concentric with multiple expansion structure gives highest noise reduction performance for lower frequency noises ranging from 600 Hz to 1000 Hz. Increase of chamber width in multiple expansion structure widens the frequency bandwidth for noise reduction and decreases the peak transmission loss.
10 CONCLUSIONS AND SUMMARY
The air path from the sound generating end is a crucial spot for passive noise reduction by means of sound energy absorption and multiple reflections. Absorption of sound takes place due to viscous dissipation in the air field caused by velocity differences between air molecules. Use of solid obstacles inside the air path promotes the sound absorption due to viscous dissipation. As the amplitude of sound wave is maximum at a quarter of its wavelength, the length of solid surface contact needs to be greater than quarter of the wavelength. Structures with continuous pores are particularly useful for accomplishing the sound absorption due to increased surface contact.
Helmholtz cavity works with principle of increasing the amplitude of sound waves near the contacting surface by creating resonance vibration. It helps in dissipation of low frequency waves with minimal length of surface contact. Another way of sound cancelling is possible by interfering a sound wave with another sound wave having 180ophase difference. The sound from same source is allowed to travel in two different paths and meet each other when a phase difference of 180o is obtained. This method is called as passive destructive interference. Creating partitions in muffler devices increases the characteristic length of sound travel that causes the increase of sound transmission loss through the muffler. The noise reduction performance of combined partitions is remarkably higher than that of individual partitions.
This study concludes that design complexity of cavity is a potential parameter affecting the performance of passive noise reduction. By creating multiple expansions along the sound carrying air path, the peak of sound transmission loss increases steeply. Increasing the width of each expansions, decreases the peak of sound transmission loss but increases the frequency bandwidth of sound reduction from lower frequency zones. By replacing 50%
width of multiple expansion structure by a single expansion chamber, the peak transmission loss of the structure decreases almost 40%. Use of Helmholtz cavity concentric to multiple expansion structure causes higher sound transmission loss in low frequency regions.
49
Multiple expansion structure introduced in the study provides a peak transmission loss of 240 decibels and completely controls the sound less than 50 decibels for medium frequency range from 1000 Hz to 3250 Hz. Use of combination of Helmholtz cavity with multiple expansion structure in a concentric arrangement, provides a peak transmission loss of 173 decibels for a frequency of 620 Hz and completely controls the sound less than 50 decibels for a frequency range between 600 Hz to 1000 Hz.
The use of additive manufacturing technologies such as FDM and PBF are suitable for manufacturing complex structures with continuous pores that uses two or more sound absorption mechanisms. The design process of noise reduction cavity structure should also account for design for additive manufacturing as discussed in this study. The figure 40 shows the strategy flow for development of additively manufacturable noise reduction cavity structures working on atmospheric conditions.
Figure 40. Strategy for development of AM based noise reduction cavity structures.
As shown in figure 40, the sound absorption elements are first chosen according to frequencies of noise to be reduced. A combination of absorption mechanisms is developed by assembling different absorption elements to design a noise reduction cavity. The design of the cavity is checked with DFAM rules and corrections are to be made. The cavity is numerically modelled by finite elements and the STL value is calculated. Based on peak
Identification of input noise properties Stratagy for selection of sound absorption elements Design of noise reduction cavity based on absorption elements
Correction of cavity design according to rules of DFAM
Acoustic analysis of cavity using FEM to calculate sound transmission loss Optimize the design for higher transmission loss and larger bandwidth of frequency
Modelling wall structure for noise reduction cavity Correction of wall structure design according to rules of DFAM
Additive manufacturing of wall structure of cavity
transmission loss and bandwidth of frequencies of higher transmission loss, the design parameters of the cavity are optimized. Wall structure for the cavity is designed according to rules of DFAM and finally manufactured using FDM or PBF process.
51
11 FURTHER STUDY
The noise reduction performance of cavity structures is to be analysed for different temperatures and fluid flow velocity depending on the specific applications. Experimental analysis of sound transmission loss for new cavity structures has to be carried in different working conditions.
Light weight wall structures as shown in figure 41 are manufacturable with wide range of innovative designs using additive manufacturing. The use of air gaps inside the walls increases the noise reduction performance in light weight constructions (Harun et al. 2012, p. 246). The future study has to be carried on structure and air borne noise reduction across different hollow wall structures.
Figure 41. Cross section of sandwich cored light weight walls (Bagsik et al. 2014, p. 697).
As shown in figure 41, the light weight wall structures have a combination of fluid and solid media. A vibro-acoustic analysis has to be carried to investigate the sound transmission loss across the structure that is made of materials with different densities.
REFERENCES
Abele, E., Stoffregen, H., Klimkeit, K., Hoche, H. & Oechsner, M. 2015a. Optimisation of process parameters for lattice structures. Rapid Prototyping Journal, 21:1. Pp. 117–127.
Abele, E., Stoffregen, H., Kniepkamp, M., Lang, S. & Hampe, M. 2015b. Selective laser melting for manufacturing of thin-walled porous elements. Journal of Materials Processing Technology, 215. Pp. 114–122.
Adam, G. & Zimmer, D. 2014. Design for Additive Manufacturing—Element transitions and aggregated structures. CIRP Journal of Manufacturing Science and Technology, 7:1. Pp.
20–28.
Adam, G. & Zimmer, D. 2015. On design for additive manufacturing: evaluating geometrical limitations. Rapid Prototyping Journal, 21:6. Pp. 662–670.
Arenas, J.P. & Crocker, M.J. 2010. Recent trends in porous sound-absorbing materials.
Sound & vibration, 44:7. Pp.12–18.
ASTM F2792-12a. 2013. Standard Terminology for Additive Manufacturing Technologies, ASTM International, West Conshohocken, PA. 3 p.
Bagsik, A., Josupeit, S., Schoeppner, V. & Klemp, E. 2014. Mechanical analysis of lightweight constructions manufactured with fused deposition modelling. 29th International Conference of the Polymer Processing Society-Conference Papers. AIP Publishing, 1593:1.
Pp. 696–701.
Barron, R. 2003. Industrial noise control and acoustics. New York: Marcel Dekker. 534 p.
Bohn, D. A. 1988. Environmental effects on the speed of sound. Journal of Audio Engineering, 36:4. Pp. 1–9.
53
Biswas, S. & Agrawal, A. 2013. Noise reduction in a large enclosure using single, dual and ensconced Helmholtz resonators. Current Science, 104:12. Pp. 1681–1691.
COMSOL Inc. COMSOL Multiphysics 5.2. 2015. Automotive muffler [Online document]
[Accessed 2 May 2016] Available at: https://www.comsol.fi/model/download/261801/
models.mph.automotive_muffler.pdf
Dassault Systems, SolidWorks Corp. 2015. Available at: http://www.solidworks.com.
European Environment Agency. 2014. Noise in Europe 2014. EEA Report No 10/2014. 20p.
[Online document] [Accessed 10 June 2016] Available at: http://www.eea.europa.eu/
publications/noise-in-europe-2014
Fuchs, H. 2013. Applied acoustics. Berlin. Springer. 593 p.
Gai, X., Xing, T., Li, X., Zhang, B. & Wang, W. 2016. Sound absorption of microperforated panel mounted with Helmholtz resonators. Applied Acoustics, 114. Pp. 260–265.
Geslain, A., Groby, J., Dazel, O., Mahasaranon, S., Horoshenkov, K. & Khan, A. 2012. An application of the Peano series expansion to predict sound propagation in materials with continuous pore stratification. The Journal of the Acoustical Society of America, 132:1. Pp.
208–215.
Godbold, O., Kang, J., Soar, R. & Buswell, R. 2007. From MPA to strategically designed absorbers using solid freeform fabrication techniques. Proceedings of the 19th International Congress on Acoustics (ICA). Madrid, Spain. Pp. 1–6.
Goldbold, O. 2008. Investigating broadband acoustic adsorption using rapid manufacturing.
Doctoral Thesis. Loughborough University. 197 p. Available at:
https://dspace.lboro.ac.uk/2134/8058
Gu, D. & Shen, Y. 2008. Processing conditions and microstructural features of porous 316L stainless steel components by DMLS. Applied Surface Science, 255:5. Pp. 1880–1887.
Hannink, M. H. C. 2007. Acoustic resonators for the reduction of sound radiation and transmission. Doctoral Thesis. Enschede: University of Twente. 170 p. Available at:
http://doc.utwente.nl/58025
Harun, Z., Awad, E., Nor, M.J.M. & Razi, M. 2012. A comparative study on the effect of air gap on sound transmission loss provided by double glazed panels. Applied Mechanics and Materials. Trans Tech Publications, 165. Pp. 242-–246.
Kannan, S. & Senthilkumaran, D. 2014. Investigating the influence of electroplating layer thickness on the tensile strength for fused deposition processed ABS thermoplastics.
International Journal of Engineering and Technology, 6:2. Pp. 1047–1052.
Klaus, T.B., Herold, S., Heuss, O. & Sabirin, C.R. 2012. Noise reduction of a sound field inside a cavity due to an adaptive Helmholtz resonator. Proceedings of international conference on noise and vibration engineering—ISMA2012, Leuven. Pp. 489–504.
Kranz, J., Herzog, D. & Emmelmann, C. 2015. Design guidelines for laser additive manufacturing of lightweight structures in TiAl6V4. Journal of Laser Applications, 27: S1.
Pp. S14001 1–16.
Kruth, J. P., Mercelis, P., Van Vaerenbergh, J. & Craeghs, T. 2007. Feedback control of selective laser melting. Proceedings of the 3rd international conference on advanced research in virtual and rapid prototyping. Pp. 521–527.
Liu, Y., Yang, Y. & Wang, D. 2016. A study on the residual stress during selective laser melting (SLM) of metallic powder. The International Journal of Advanced Manufacturing Technology. Pp. 1–10.
Miranda, J. G. O. & Duarte, M. A. V. 2015. Active Noise Control in Ducts: Sensor Positioning Using FRF and FEM Analysis. International Journal of Engineering and Innovative Technology, 4:7. Pp 20–24.
