A new mount with moving-magnet type electromagnetic actuator for naval shipboard equipment

Transcription

1 csnak, 2015 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 htt://d.doi.org/ /ijnaoe ISSN: , eissn: A new mount with moving-magnet tye electromagnetic actuator for naval shioard equiment Yun-Ho Shin 1, Seok-Jun Moon 1, Jeong-Il Kwon 1, Woo-Jin Jung 2 and Jae-Jin Jeon 2 1 System Dynamics Research Deartment, Korea Institute of Machinery and Materials, Daejeon, Korea 2 Agency for Defense Develoment, Changwon, Korea ABSTRACT: This study is roosed a new hyrid mount having a moving-magnet tye electromagnetic actuator to reduce the viration transmitted from naval shioard equiment to the structure of the shi s hull. Otimal design secifications are determined through eerimental analysis. The detailed design of the hyrid mount is determined through several design stes with electromagnetic numerical analysis using Mawell Software(S/W). The hyrid mount that comines a ruer mount and an electromagnetic actuator has a fail-safe function for shock resistance. The mount is faricated and tested using a universal testing machine to evaluate the design secifications. Finally, numerical simulation of the hyrid mount is erformed to confirm control erformance and alicaility. KEY WORDS: Viration isolation; Hyrid mount; Electromagnetic actuator; Naval shioard equiment. INTRODUCTION Increasing demands for greater recision in a wide range of industrial equiment and the needs for defense equiment to have etter detection-avoiding ailities have caused researchers to focus more on imroving the erformance of mounts attached to the ottom of such equiments. Widely used eamles include assive mounts, such as air mounts, ruer mounts, and wire roe mounts. Overall industrial demands for hyrid mounts are on the increase (Moon et al., 2013). A hyrid mount can eert higher erformance than a assive mount y allowing it to e comined with an (semi) active element. Though there are many tyes of hyrid mounts availale that can e comined with (semi) active actuators, they need to modify the assive art or redesign it (Moon et al., 2008; 2011; Kim et al., 2011). But most of researches are focused on the active art design as neglecting the assive art erformance or weight of ojective equiment (Moon et al., 2008; Kim, 2001). This aer uilds on the wider literature on this toic y discussing only the hyrid mount in connection with an electromagnetic actuator. Electromagnetic actuators have lower actuating forces for their volumes than other actuators. They can e alied for low frequency ranges, ut generate greater dislacement. There is a wealth of studies which aim to hel meet tightening regulations on the viration levels involved in recision roduction and test equiment in the electricity and electronics sectors. One such study is into a mount system that comines a assive air sring and an electromagnetic actuator (Moon et al., 2011; Kim et al., 2013). In the research, an electromagnetic actuator is used as a means to reduce the resonse caused y having an air sring in low resonant frequency ranges. An electromagnetic actuator is therefore used in a frequency range elow 10 Hz, and a magnetic earing tye of actuator has een develoed. In the automotive sector, studies have een reorted aout hyrid mounts that Corresonding author: Seok-Jun Moon, sjmoon@kimm.re.kr This is an Oen-Access article distriuted under the terms of the Creative Commons Attriution Non-Commercial License (htt://creativecommons.org/licenses/y-nc/3.0) which ermits unrestricted non-commercial use, distriution, and reroduction in any medium, rovided the original work is roerly cited.

2 2 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 comine ruer mounts and electromagnetic actuators intended to ensure a comfortale ride for assengers y reventing the viration from the engine eing carried to the ody (Lee et al., 2009). The majority of such mounts are designed to oerate through the actuator, ased on the engine revolutionary seed. This aer discusses a hyrid mount that comines a assive ruer mount and an electromagnetic actuator that can e alied to naval shioard equiment. Active arts are added to the erformance guaranteed assive ruer mount that are frequently alied to a naval vessel without its modification, differentiating this tye of mount from other tyes of active mounts. The hyrid mount eamined in this aer requires less ower to suort a static ayload (weight of ojective equiment) y the assive arts, without any adverse effects to the mount isolation erformance. Furthermore, it is easy to install or add to a location where a ruer mount has already een installed. This study is roosed a new tye of hyrid mount caale of suressing the viration from equiment and reventing it from eing transferred to the ody of a naval shi. As such, it rotects equiment from eternal shock since the air ga etween the mover and the stator of an electromagnetic actuator enales it to oerate on a non-contact asis, making it ideal for shock-resistant mounts for naval shioard equiment.. An electromagnetic actuator can e either a moving-coil tye or a moving-magnet tye. With research into a hyrid mount incurorating the former already reorted (Shin et al., 2013), this aer discusses a hyrid mount incororating a moving-magnet ased electromagnetic actuator. First, the design secifications for the electromagnetic actuator were determined, and the design was conducted in light of various design factors, including shock resistance. The hyrid mount designed was manufactured and sujected to eeriments and simulation to measure its erformance. DETERMINATION OF DESIGN SPECIFICATIONS The design secifications of the hyrid mount resented in this aer were determined ased on the viration measurements of the system ("target system") that the mount will e alied to. Fig. 1 shows an illustration of the target system consisting of a motor, a um, and four assive ruer mounts. The water is circulated through ie system directly connected y a um. In fact, there are many comonents connected to the target system, which are not included in the figure for demonstration urose. Accelerometers were installed at the to and ottom of each of the mounts (locations 1, 2, 3, and 4) to identify how viration from the motor and the um is delivered through the mounts to the ground. In order to measure viration in the vertical (heave) and the rotating directions (roll/itch), several accelerometers were installed for the z-direction. Fig. 1 Schematic diagram for the target system. The measurement in the vertical direction otained at mount 1 (location 1) for a rated revolution of 1,600 rm is given in Fig. 2. Fig. 2(a) illustrates acceleration measured at the to and the ottom of the mount at a frequency range elow 300 Hz. The measurement reveals that eaks are reached at 26.7 Hz, the first rotational comonent, and at 53.3 Hz, the second rotational comonent. It also reveals that the assive ruer mount can reduce viration transmitted. The same data can e as shown in Fig. 2() y a one-third octave and and through comarison with the standard for viration criteria (MIL-STD-740-2, 1986).

3 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 3 Viration at the to eceeds tolerance at some range while viration at the ottom eceeds the tolerance at the first rotational frequency. While it is not easy to estimate the actuation of the target system y using acceleration signal measured for it, it is ossile to estimate the force transmitted towards the ground through the mount if the material information on the mount is known. In other words, the transmitted force ( F T ) can e estimated from the asolute acceleration ( ) measured at the to and the ottom of the mount using the following equation. ( ) ( ) F = c + k (1) T r e r e where c r, kr reresent the linear daming and the linear stiffness of the ruer mount resectively, and e, reresent equiment and ase resectively, meaning the to and the ottom of the mount. Velocity and dislacement are calculated from the measurements of acceleration signals as integrating them numerically, and Eq. (1) is used to calculate the transmitted force. The transmitted force is given in Fig. 3(a) and can e reresented in the frequency domain as shown in Fig. 3(). Assuming that any frequency comonent with a difference of 20 db or greater from the maimum transmitted force measured is small enough to e removed from the frequency comonents that need to e controlled, interested frequency comonents and the oerating force required for each of them are as shown in Tale 1. For the conservative design of the actuator, its maimum oerating force required and maimum drive dislacement were determined at 120 N and ±1 mm resectively, as ased on the transmitted force in the time domain. These, along with constraints for sace, are defined as the required design secifications of an electromagnetic actuator. (a) Linear scale. () 1/3 octave scale. Fig. 2 Acceleration signals at location 1. (a) In time domain. () In frequency domain. Fig. 3 Transmitted force through the mount at location 1.

4 4 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 Tale 1 Required design secification of an electromagnetic actuator. Domain Frequency domain Time domain Required force / dislacement 41.5 N 26.5 Hz 3.3 N 53.3 Hz Ma. 120 N eak (force) Ma. ±1 mm (dislacement) DESIGN OF A HYBRID MOUNT The hyrid mount resented in this aer incororates a assive ruer mount locally roduced ased uon the U.S. Navy's standard mount. This mount named 6K900 has a resonant frequency of 6 Hz at allowale static load of aout 400 kg (see Fig. 4) and is currently used as a naval shioard mount. The moving magnet ased electromagnetic actuator was designed to e attached to the ottom of the assive mount, and any change in the structure that can affect the erformance of the assive mount was not considered. In terms of the design of the moving magnet ased electromagnetic actuator, the ermanent magnet was laced inside as a mover while the coil was laced outside as a stator. Fig. 5(a) shows a method of design where two ermanent magnets with different olarities are laced inside to imrove erformance, while Fig. 5() illustrates a method to reduce magnetic flu leakage at the coil surrounding the magnet. The candidate designs were sujected to electromagnetic analysis y means of Mawell, a commercial finite element solution Software(S/W). For a comarison of the characteristics with the moving-coil tye, an analysis was conducted for the design shown in Fig. 3(a) of Shin et al., (2013). Fig. 4 Drawings of the assive mount, 6K900 (Suer Century Co., 2011). (a) Tye 1. () Tye 2. Fig. 5 First ste design for a moving-magnet electromagnetic actuator.

5 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 5 The candidate designs were eamined in terms of flu density and ack electromotive force (ack EMF). They were found to e indirectly and directly roortional with generated force. Fig. 6 shows art of the results from the Mawell analysis, namely the magnetic flu density at the air ga, and Fig. 7 shows the characteristics of ack EMF which is the generated voltage as assigning 1 m/s velocity to a moving art of each candidate. The unit of ack EMF, Vs/m, is equivalent with the generated force er unit current, N/A. Fig. 6 suggests that the moving-coil tye with a smaller change in magnetic density at the air ga has etter control erformance than the moving magnet tye. It also suggests that the moving coil tye has etter current resonse ecause it has relatively lower resistance and inductance. Fig. 7 shows that the moving-magnet tye (tye 1) has greater ack EMF, which suggests that it has etter force characteristics. In contrast, the moving-magnet tye (tye 2) is redicted to have stale force characteristics over a wide oerating range. 1.0 Moving Coil Tye Moving Magnet Tye1 Moving Magnet Tye Moving Coil Tye Moving Magnet Tye1 Moving Magnet Tye2 Airga Fludensity [T] Back EMF [Vs/m] Distance [mm] Fig. 6 Flu density at air ga Time [ms] Fig. 7 Back EMF coefficient. In order to imlement the moving-magnet tye, it is imerative to address technological challenges associated with the structure suorting the ermanent magnet and find ways to imrove its control erformance. Fig. 8 shows a concet that can ossily meet such requirements. For tye 1, some of the magnets located outside were moved to inside with more sace etween them. For tye 2, the ositions of the coil and the ermanent magnets were switched to determine the shock on its erformance. Results from electromagnetic analysis are shown in Figs. 9, 10 and 11, each of which indicates the magnetic flu, the ack EMF constant and the detent force at the air ga. The results were comared with those shown in Fig. 3() of the reference (Shin et al., 2013). Fig. 9 shows that the magnetic flu density at the air ga is much more stale than in Fig. 6. The moving-magnet tye, in articular, shows a significant imrovement, which seems likely to increase control erformance. Fig. 10 shows that the ack EMF constant for moving magnet tye 1 more than douled. It is therefore resumed that its oerating force likewise more than douled. Fig. 11 shows the detent force that may distur oeration: the lower the value, the greater the erformance ecomes. This suggests that moving magnet tye 2, desite its smaller size, has etter characteristics of dynamic oerating force. Fig. 12 shows the oerating force, which is ased on inut current calculated at the center of the electromagnetic actuator. Conservatively, moving-magnet tye 1 requires an inut current of 2 A to gain oerating force of 200 N while tye 2 requires large current of 5 A or more to gain the same force. (a) Tye 1. () Tye 2. Fig. 8 Second ste design for a moving-magnet electromagnetic actuator.

6 6 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 The design roosed in Fig. 8 lacks the fail-safe caaility required for a naval mount. This means that uon receiving an eternal shock, the electromagnetic actuator may disrut the deformation of the assive ruer mount. The design was therefore revised to avoid any imairment of the assive mount's shock resistance due to the electromagnetic actuator. Based on the design in Fig. 8() which features oor static force characteristics ut good dynamic static force characteristics, the actuator was redesigned, as shown in Fig. 13. The redesigned actuator can accommodate deformation of u to 5 cm eected from shock load and has its ermanent magnet the mover located inward. The outward location of the coil segment was found to reduce force generated from the actuator. The height of the entire hyrid mount was not twice as high as the assive mount and its structure was designed to e simle enough for ease of roduction. Its air ga is 2 mm, its coil is wraed with 270 turns, and its weight is aout 10 kg Moving Coil Tye Moving Magnet Tye1 Moving Magnet Tye2 Back EMF [Vs/m] Time [ms] Fig. 9 Flu density at air ga. Fig. 10 Back EMF coefficient Moving Coil Tye Moving Magnet Tye1 Moving Magnet Tye2 300 Detent Force [N] Time [ms] Fig. 11 Detent force. Fig. 12 Static force at center osition. Fig. 13 Third ste design for an electromagnetic actuator.

7 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 7 Fig. 14 shows results from an analysis using Mawell for one cross-section only. They give an insight into the magnetic flu density and force created at the rotor and the stator when currents of 1 A, 2 A and 3 A are alied to the coil. The magnetic flu density increased in roortion to the magnitude of the current alied, and the increased density at the air ga caused the actuator to have additional ower of 72 N, 145 N and 219 N over time. There is the ossiility that a change in the magnitude of the static deflection may re-define the relative locations of the coil and the ermanent magnet, affecting the ower of the actuator. The shock of a change in the weight of equiment mounted was, therefore, elored, and re-interretation was conducted for a change of -4 mm to 7 mm in deflection at design load of 250 kg. Results from the interretation are revealed in Fig. 15. It was found that the degree of deflection affects the magnitude of ower created. In this study, we eect that, ased on the characteristics of the assive mount, the change in the static deflection will e less than ±2 mm and the change in the in the force will e around 20 %. (a) 1 A. () 2 A. (c) 3 A. Fig. 14 Electromagnetic analysis results. Fig. 15 Electromagnetic force w.r.t. static deflection.

8 8 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 The design secifications and arameters finally derived are summarized in Tale 2. The draft design for a hyrid mount incororating an electromagnetic actuator ased on the said secifications and arameters is given in Fig. 16. The lack and the lue comined reresent the ruer mount: the lack reresents steel; and the lack, ruer. Tale 2 Design secification of an electromagnetic actuator. Item Electrical air ga Mechanical air ga Wire secification Weight Resistance Inductance Value 5 mm 2 mm 1.3 mm 270 turns a. 10 kg 2.33 Ω 0.22 H Fig. 16 Hyrid mount with the electromagnetic actuator. FABRICATION AND PERFORMANCE EVALUATION Farication of the hyrid mount As mentioned in the revious section, the 6K900, which delivers erformance equivalent to that of a U.S. Naval standard mount, was emloyed as the assive ruer mount. The moving magnet ased electromagnetic actuator resented in the revious section was manufactured in-house. The metal used to create a magnetic circuit was ferrite chrome steel, or SS430, while the ermanent magnet was made from N42H. The magnet was uilt y manufacturing 20 ieces of it through discretization and assemling them into a ring shae. This was uilding a ring-shaed magnet of a single iece at once was unfeasile. The coil is of self-ounded tye and suffers over time from the artial melting of the sheath covering it, once the current egins to run through it. The diameter was finally determined at 1.3 mm in light of the current alied. All other arts ecet the one necessary for the creation of a magnetic circuit were made from non-magnetic stainless steel to avoid magnetic flu leakage. Fig. 17 shows the stator and the rotor manufactured and the hyrid mount. The mover consists of a ermanent magnet and other comonents. A closer look at the hotos in Fig. 17() will reveal that white ieces of ermanent magnet are connected with each other outside the stator. The stator shown in Fig. 17(a) consists of a coil and other comonents. Fig. 17(c) shows an electromagnetic actuator consisting of a stator and a mover, as attached to the ottom of the assive ruer mount. It shows that the actuator is connected to an installation jig.

9 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 9 (a) Static art. () Mover art. (c) Assemled hyrid mount. Fig. 17 Assemled hyrid mount and key element arts. General erformance eeriment A Universal Testing Machine (UTM) was used to test the erformance of the electromagnetic actuator uilt to the design secifications resented earlier. As shown in Fig. 18, the hyrid mount was laced in the UTM, and a force sensor was attached to the connection at the to of the mount. An eeriment to determine the equiment's erformance was configured, as detailed in Fig. 19. The UTM alied a force of 250 kg equivalent to the weight of equiment at the to of the mount. A DC driver was set u to suly current to the electromagnetic actuator and was owered through connection to a stale ower suly. A ulse generator (B&K) was used to control the amlitude and waveform of the electric current. A force sensor measured the force generated from the mount and transmitted the measured value to the comuter for storage. The stored data were analyzed to evaluate the erformance of the hyrid mount. Fig. 18 Eerimental set-u. Fig. 19 Eerimental configuration (Shin et al., 2013).

10 10 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 A sinusoidal wave of a single frequency was alied to measure the force generated from the actuator. The eeriment was conducted at an interval of 10 Hz for a frequency range of 10 Hz to 100 Hz and at an interval of 25 Hz for the 100 Hz to 300 Hz range. Fig. 20 shows the results from an eeriment where a current of 1 A rms was alied. The results reveal that the force decreased dramatically as the frequency increased, efore stailizing in the over 100 Hz range. The force increased at a stronger current, ut not in linear roortions. When the current was 1 A rms, the force was a maimum 1,200 N at 10 Hz, while it was 80 N at 300 Hz. Tale 3 rovides a summary of the eeriment results, ased searately on the design secifications in Tale 2 and on a hyrid mount incororating an eisting moving coil ased electromagnetic actuator. As shown in Tale 3, the required design secifications are met. Inut Current: 1.41 A Peak Force [N] Frequency [Hz] Fig. 20 Eerimental results. Tale 3 Eerimental results of an electromagnetic actuator. Domain Moving magnet tye Moving coil tye Remark Electrical air ga 930 N eak 405 N 26.5 Hz Mechanical air ga 548 N eak 124 N 53.3 Hz Wire secification 1,200 N eak 900 N eak Ma. value Numerical simulation of control erformance A numerical simulation was conducted to evaluate the control erformance of the hyrid mount. The target system shown in Fig. 1 can e riefly modeled using a degree-of-freedom system, as shown in Fig. 21. The equations of motion are given elow, as Eqs. (2) and (3). Fig. 21 Modeling for the target system.

11 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 11 M ( ) + k ( ) = F F0 + c (2) r r a + r ( ) + c + k ( ) + k = Fa m + c (3) where, reresent dislacement of the equiment ( M ) and dislacement at the ottom of the mount( m ), resectively. M, cr, k reresent the mass of equiment mounted and the linear daming coefficient and the linear stiffness coefficient of r the ruer comonent of the hyrid mount, resectively. m, c, k reresent the mass, the attenuation coefficient and stiffness coefficient of the installation jig for the hyrid mount, resectively. F a, F reresent the control generated from the 0 electromagnetic actuator of the hyrid mount and the eciting force generated from the equiment mounted, resectively. The equations of motion for an electromagnetic actuator are given elow. r Fa = kti (4) T ( ) Li = V k (5) where k T, L, R reresent a ack EMF constant, inductance and resistance, resectively, while i, V reresent inut current of the coil and voltage, resectively. The state variales and equation are summarized in Eqs. (6) and (7) and the values of arameters in the mathematical model are descried in Tale 4. X = i [ ] F = F V o T / / / / / kr M cr M kr M cr M kt M 1/ M X = X F ( kr + k) ( cr + c) k / / / r m cr m kt m 0 0 m m 0 1/ 0 / 0 / / L kt L kt L R L T (6) (7) Tale 4 Parameter secifications in a mathematical model. Parameters Values Parameters Values M 250 kg c Ns/m m 9 kg k T 70 N/A k r N/m L 0.20 H c r Ns/m R 1.17 Ohm k N/m The first numerical simulation involved considering when the mounted equiment would e oerational at 1,600 rm. The control erformance measured when the electromagnetic actuator was disaled. In addition, the oints of activation were comared. Comarison was made of the alicaility of a PD control and a Filtered-X LMS (Widow and Stearns, 1985; Hakansson et al., 1998)when the actuator was on. The control was designed to reduce the 26.7 Hz (1,600 rm) comonent, the rimary frequency. The results of the simulation are summarized in Tale 5. The frequency domain shows the aility of the 26.7 Hz comonent to reduce viration, while the time domain shows RMS values for all comonents elow 300 Hz. When the electromagnetic actuator is off (no control), the assive ruer mount reduces viration y 6 db efore transmission.

12 12 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 It was found that when the actuator is on (control), frequency comonents can e reduced y 23 db to 83 db, while they can e reduced y 22 db to 44 db in the time domain. The Filtered-X LMS control rovides nearly twice the level of control erformance than the PD control. Tale 5 First simulation results for control erformance of the hyrid mount. Control Control logic Variale 26.7 Hz Time domain Control off Control on None PD control Filtered-X LMS control m 2 /Hz (eak) m rms m 2 /Hz (eak) m rms Ratio ( / ) -6 db -6 db m 2 /Hz (eak) m rms m 2 /Hz (eak) m rms Ratio ( / ) -23 db -22 db m 2 /Hz (eak) m rms m 2 /Hz (eak) m rms Ratio ( / ) -83 db -44 db (a) Actuator voltage and current. () Generated actuator force. (c) Time domain resonse. (d) Frequency domain resonse. Fig. 22 Passive 2-DOF system (Actuator off, Target: Reduction of Base Viration, ).

13 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 13 The eciting force of the target system, which consists of a motor and a um, includes frequency and harmonic comonents. For this reason, simulation was conducted for a second time to determine the control erformance of the harmonic comonent. Figs. 22 and 23 and Tale 6 show frequencies corresonding to 1,600 rm and an eciting force that includes second and third harmonic comonents. The control emloyed Filtered-X LMS and summaries are variously made of scenarios when one frequency is considered and when three are considered. The control intended to reduce one frequency shows a reduction of 32 db in the time domain, ut fails to show notale control erformance for the second and third harmonic comonents. This is in contrast with the control ased on three frequency comonents, which shows a reduction of over 60 db for each harmonic comonent. Interestingly, the force generated in the assive simulation is due to the relative movement etween and, as shown y Fig. 22. The movement induces the current and it generates a force without inut command voltage. However, this generated force is relatively small comared with the control force shown in Fig. 23 and is negligile in terms of the simulation and realization of the designed actuator. The required inut current and voltage to suress the target ase viration is 12 V eak, 0.4 A eak and the force is low 25 N eak. The control force is a small dynamic force designed to reduce the ase viration, which could otherwise induce or ecite the other equiment resonance or critical frequencies. This is ecause the assive ruer mount suorts a static load (the weight of equiment) and suresses equiment viration. (a) Actuator voltage and current. () Generated actuator force. (c) Time domain resonse. (d) Frequency domain resonse. Fig. 23 Active 2-DOF system (Actuator on, Target: Reduction of Base Viration, ).

14 14 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 Tale 6 Second simulation results for control erformance of the hyrid mount. Control Control logic Variale 26.7 Hz (m 2 /Hz (eak) ) 53.3 Hz (m 2 /Hz (eak) ) 80.0 Hz (m 2 /Hz (eak) ) Time domain (m rms ) Control off Control on None Filtered-X LMS Control (1 frequency comonent) Filtered-X LMS Control (3 frequency comonents) Ratio ( / ) -6 db 0.4 db -2 db -6 db Ratio ( / ) -83 db 0.4 db -2 db -32 db Ratio ( / ) -85 db -60 db -60 db -44 db CONCLUSIONS This aer discusses the develoment of a new hyrid mount caale of reducing structure-orn noise caused y naval shioard equiment. Building on revious research y the authors evaluating the moving-coil electromagnetic actuator, this aer eamines the moving magnet-ased actuator. Key findings from the study are following: 1) The new hyrid mount comines a roven assive ruer mount and a moving magnet-ased electromagnetic actuator. It is ideal for reducing viration at a low frequency range (under 300 Hz) and has fail-safe caaility that can rotect the equiment at the to of the mount from eternal shock. 2) The moving magnet ased electromagnetic actuator is designed to fully meet regulatory guidelines on coils, ermanent magnets and magnetic circuits. 3) A hyrid mount was roduced ased on the design and thoroughly tested using a hydraulic universal testing machine to determine its erformance. It was found that the moving-magnet tye has greater oerating force and comlies with the design secifications in the frequency range under oservation. 4) Numerical simulation was conducted to evaluate the control erformance of the hyrid mount, and its alicaility was confirmed. Significantly, the alicaility of the Filtered-X LMS algorithm to harmonic comonents was demonstrated. 5) Performance evaluation of a mount system that consists of the four mounts is necessary to determine the overall erformance of the hyrid mount, and further research is required in this area. ACKNOWLEDGEMENTS The research descried in this aer was funded y the Agency for Defense Develoment (Project No.: , Contract No.: UD11036DD). REFERENCES Hakansson, L., Clasesson, I. and Sturesson, P.O., Adative feedack control of machine-tool viration ased on the filtered-x LMS algorithm. International Journal of Low Frequency Noise, Viration and Active Control, 17(4), Kim, H.T., Kim, C.H., Kang, S.B., Moon, S.J. and Lee, G.S., A hyrid structure of dual stators and a neumatic sring for resonance control in an air mount. Journal of Electromagnetic Analysis and Alications, 5, Kim, U.Y., Design and alication of active damer using electrodynamic actuator. Master Degree Thesis. KAIST.

15 Int. J. Nav. Archit. Ocean Eng. (2015) 7:41~55 15 Kim, W.G., Kim, Y.S., Lee, W.C. and Kim. K.H., Performance otimization of electromagnetic active engine mount. Proceedings of the Korean Society for Noise and Viration Engineering Conference, 1, Lee, B.H. and Lee, J.W., Model ased feed-forward control of electromagnetic tye active control engine-mount system. Journal of Sound and Viration, 323, Military Secification MIL-STD-740-2, Structureorne viratory acceleration measurements and accetance criteria of shioard equiment. Washington D.C.: Deartment of Defense, Availale at: <htt://everysec.com/mil-std/ MIL-STD /MIL-STD-740-2_10380/> [Accessed Novemer 2014]. Moon, S.J., Choi, S.M., Nguyen, V.Q., Oh, J.S., Choi, S.B., Chung, J.H., Kwon, J.I. and Jung, W.J., An inertia-tye hyrid mount that comines a ruer mount and a iezostack actuator for naval shioard equiment. International Journal of Naval Architecture and Ocean Engineering, 5, Moon, S.J., Ji, Y.J., Yoon, J.S., Choi, S.B., Lee H.Y., Kim, J.H. and Jung, W.J., A study on develoment of an active hyrid mount for naval shis. Journal of the Society of Naval Architects of Korea, 45(3), Moon, S.J., Park, S.H., Jeong, J.A., Huh, Y.C., Kim, C.H. and Choi, S.M., A study on the develoment of a hyrid electromagnetic actuator against microviration. Transactions of the Korean Society for Noise and Viration Engineering, 21(5), Shin, Y.H., Moon, S.J., Cho, H.Y., Jung, W.J., Jeon, J.J. and Won, M.C., A new mount with moving-coil tye electromagnetic actuator for naval shioard equiment. Transactions of the Korean Society for Noise and Viration Engineering, 23(10), Suer Century, [online] Resilient mounts. Availale at: <htt:// [Accessed Novemer 2014]. Widrow, B. and Stearns, S.D., Adative signal rocessing. Englewood Cliffs, NJ: Prentice-Hall.