Experimental Validation of a Suspension Rig for Analyzing Road-induced Noise

Transcription

1 Expermental Valdaton of a Suspenson Rg for Analyzng Road-nduced Nose Dongwoo Mn 1, Jun-Gu Km 2, Davd P Song 3, Yunchang Lee 4, Yeon June Kang 5, Kang Duc Ih 6 1,2,3,4,5 Seoul Natonal Unversty, Republc of Korea 6 Hyunda Motor Group, Republc of Korea ABSTRACT As an nterest of road-nduced nose wthn a vehcle ncreases, the operatonal deflecton shape (ODS) analyss has been used to dentfy a suspenson dynamc behavor n the actual operatonal condton and to establsh countermeasures to the problem. In ths study, ODS comparson between the baselne vehcle and the suspenson rg was conducted for analyzng a suspenson road-nduced nose performance. Before the ODS measurement, the major contrbuton path of structure-borne road nose transmtted through a suspenson was verfed by usng the Transfer Path Analyss (TPA) method. In order to evaluate the effect of the major moton of a suspenson on the target pont, senstvty analyss was also performed to the target by usng the measured response data of ODS. Furthermore, the comparson of the evaluaton n between the baselne vehcle and the suspenson rg was verfed on the bass of the TPA and senstvty analyss results. Through the comparson results, the usefulness and valdty of usng the suspenson rg were nvestgated n analyzng operatonal dynamc behavor for road-nduced nose. Keywords: Operatonal Deflecton Shape, Senstvty Analyss, Suspenson Rg 1. INTRODUCTION Road-nduced nose has become a major nose, vbraton, and harshness (NVH) problem wthn a vehcle because of ncreasng nterests n electrcal vehcles and many advances n vehcle engne nose reducton. Road-nduced nose defnes structure-borne nose, ncludng vbraton, harshness, and low-frequency nose regon. Thus, road nose affects the rde comfort of drvers and passengers as well as the vehcle nteror nose. Structure-borne nose s transmtted from the road to the vehcle nteror through tre and vehcle suspensons. Therefore, reducng the nose by usng an acoustc absorbent and nsulaton materal s dffcult. A structural approach s requred to reduce road nose. Accordngly, transfer path analyss (TPA) s the most popular and powerful method used to analyze the structure-borne nose. The TPA s used to verfy the energy transfer contrbutons of each selected path and propose whch path should be modfed to reduce the structure-borne nose. In the road nose, the TPA s manly appled to suspenson mount bushes connected to the vehcle body mount and lns of mult-ln suspenson. The TPA s wdely used to dentfy the major contrbuton path wthn the structure. However, t s mpossble to analyze the path s dynamc behavor through the TPA method. The dynamc behavor of the whole structure and the nterested moton that seems to nfluence the major contrbuton path should be checed to optmze and modfy the path. Thus, the operatonal deflecton shape has been used to mprove a suspenson s road nose performance. Road nose analyss tools, such as TPA and operatng deflecton shape (ODS), have been appled to a baselne vehcle to reduce vehcle road nose, whle the full matrx nverson method has been used to calculate the operatonal body mount force. However, the latter cannot accurately measure the operatonal mount force. Therefore, Davd et al. proposed a new road nose analyss method by usng a specal 1 pengundw@snu.ac.r 2 junguray@snu.ac.r 3 dsong@snu.ac.r 4 pjuny7@snu.ac.r 5 yeonjune@snu.ac.r 6 baramsolee@hyunda.com 5337

2 suspenson rg. The suspenson rg drectly measures the force and accurately uses the drect force method for structure-borne road nose TPA [1]. In ths study, the ODS analyss method s appled to the suspenson rg to analyze the road nose wthn a vehcle by dentfyng the operatonal dynamc behavor of a suspenson. A comparson between the suspenson ODS of the suspenson rg and the baselne vehcle verfes whether or not the dynamc behavor of a suspenson n the rg represents that of a suspenson n the baselne vehcle. Moreover, the senstvty analyss s used to verfy the specal rg because t s calculated based on the measured acceleraton data [2]. Applyng the ODS and senstvty analyses to the suspenson rg maes t possble to evaluate the road nose caused by only a suspenson and analyze the structural effects of a vehcle suspenson on the road nose. 2. THEORY 2.1 ODS The ODS s defned as a forced moton of the nterested ponts on a structure when the structure s n operaton. In other words, the relatve dsplacements of the ponts form an ODS, and an nformaton, such as ampltude and phase, s ncluded. The ODS analyss ams to dentfy a system s dynamc behavor n the actual operatng condton so that t wll be possble to establsh a countermeasure to the system s nose and vbraton problem by checng the most movng part at the nterested frequency range. Ths analyss requres the measurement of frequency response functons (FRFs) or spectral transmssblty functon (acceleraton/acceleraton transmssblty) to defne a structure s ODS. However, measurng the exctaton force n an operatng machnery case can be dffcult. The FRF measurement also requres that all responses caused by the exctaton force are smultaneously measured. Obtanng the response data of many selected ponts at once s dffcult. Therefore, transmssblty s used n many cases to defne the ODS nstead of the FRF, especally n evaluatng a suspenson s road-nduced nose. The spectral transmssblty functon can be expressed as follows [3]: [ T( a)] { X( a)} /{ X( a)} [ H( a)] { F} / [ H( a)] { F} (1) j j j where and j are the ndces of the response and reference ponts on the structure, respectvely. { X ( a)} refers to the response data measured by the selected ponts, and [ H ( a)] s the FRF. Thus, the rght-hand sde explans that the transmssblty functon can be measured by the FRF and the exctaton force functons when the exctaton force s measurable. However, the response of all the selected ponts cannot be smultaneously obtaned. 2.2 Desgn Senstvty Analyss The desgn senstvty analyss s used n ths paper to verfy the ODS comparson by quantfyng the measured responses of a suspenson. Km et al. [2] proposed a new desgn senstvty analyss method usng measured acceleraton data to easly analyze the structure senstvty. Beng calculated wth the measured data for the ODS, the method can quantfy the dynamc behavor of a suspenson n operaton whle the ODS s smultaneously measured. The transmssblty functon T n a structure s expressed by: ro T () (2) r where r o s the response of reference node 0, and r s the response of node of a structure. The desgn senstvty analyss s then calculated by: dro r (3) dr v where s the absolute value of a complex number; z s a small desgn modfcaton; and V s the magntude of the transmssblty functon T. The normalzed senstvty ndex S s expressed by: r v S(%) (3) n r v

3 3. ODS MEASUREMENT An experment was conducted on a chasss dynamometer n a sem-anechoc chamber to replace the real vehcle operatng condton and measure the ODS n the baselne vehcle and the suspenson rg. The operatng condton of the chasss dynamometer was 20 m/h to 120 m/h run up and run down on the patched dynamometer roller, whch represented a real rough road condton. The suspenson was a McPherson-type suspenson. Only one suspenson was used for both the baselne vehcle and suspenson rg tests to rase experment precson. Both experments were conducted usng the same tre wth 34 ps ar pressure and the same rde heght to load the same weght on the suspenson. Fgure 1 Expermental setup of the suspenson rg and baselne vehcle Fgure 2 shows the suspenson experment setup to measure the ODS. The used accelerometer sensor was a PCB 356A15 ICP tr-axal accelerometer sensor. The accelerometers were nstalled on the left and rght sdes of a lower control arm and the whole sub-frame. A total of 24 accelerometer sensors were nstalled on the left and rght lower control arms, respectvely, whle 35 accelerometers were nstalled on the sub-frame. The accelerometers were placed n three rows to dentfy the lateral and longtudnal bendng motons. Fgure 2 Installaton of the accelerometer sensors at the lower control arm and the sub-frame Fgure 3 Geometry and node numbers of the lower control arm and the sub-frame 5339

4 4. RESULTS The nucle acceleratons of the two tests were compared to valdate that both experments (.e., suspenson rg and baselne vehcle tests) were conducted n the same operatonal condton. Fgure 4 shows the comparson of the nucle acceleraton spectrum data of the experments n the 0 to 50 frequency range. The red straght lne s the nucle acceleraton of the suspenson rg, whereas the blue dashed lne s that of the baselne vehcle. The nucle acceleratons of the x, y, and z drectons are very smlar, except for the data over 35 n the x drecton. However, the dfference s not a bg problem because the exctatons of the y and z drectons are bgger and domnant over the 35 frequency range than that of the x drecton. Therefore, t s verfed that the data of both experments were measured n the same operatonal condton and had the same nput exctaton. Fgure 4 Knucle acceleraton data comparson: Left and rght sdes (the red lne s the data measured on the suspenson rg, whereas the blue lne s that on the baselne vehcle) 4.1 ODS Comparson The left and rght lower control arms were frst compared by dentfyng each ODS. The tme data-processng results n the spectrum data wth dfferent magntudes dependng on the phase reference pont. Therefore, the closest pont to the nucle was selected as the spectrum phase reference pont to calculate the acceleraton spectrum data under equal condton n both experments. Fgure 5 shows the envelope of the spectrum data of the two experments. The two spectra had the same pea ponts although the data magntudes were dfferent. Frequency peas around 8, 12, and 225 Hz were manly consdered n ths paper because the TPA results dentfed those as the major peas affectng the vehcle NVH problem. Fgure 5 Spectrum envelope (maxmum data values) of the rg and baselne vehcle tests 5340

5 Fgure 6 shows the measured ODS of the suspenson rg and baselne vehcle tests at 86 Hz, 12, and 225 Hz. As mentoned earler, the domnant dynamc motons of the component were smlar even though the two ODS data had a dfference n magntude (e.g., the bendng moton from the axs passng through the hole n the lower control arm at 226 Hz). When locally evaluated, the dynamc moton at the sde ponts was regarded as the same although the two sde ponts connected to the nucle and the sub-frame moved up and down wth dfferent phases. The ODS of the two experments at 126 Hz also had smlar motons. The lower control arms of each test domnantly moved to the x and y drectons rather than at the z drecton n comparson to the shapes at 86 Hz and 225 Hz. Lastly, the ODS at 86 Hz had bendng motons at the broad secton, whch turned nto the left sde connected to the nucle n the z drecton. Fgure 7 shows the sub-frame ODS comparson at 225 Hz. Both of the ODS had a torsonal moton based on the y axs. The rg test moton was exaggerated for better understandng although the suspenson rg test had a smaller sub-frame dynamc moton than the baselne test. Knucle Subframe Fgure 6 Comparson of the lower control arm ODS measured on the suspenson rg and baselne vehcle: top 225 Hz; mddle 126 Hz, and bottom 86 Hz 5341

6 Fgure 7 Comparson of the subframe ODSs measured on the suspenson rg and baselne vehcle at 225 Hz 4.2 Desgn Senstvty Analyss The desgn senstvty analyss could be calculated by usng Eq. (2) wth the acceleraton data measured to evaluate the ODS. The senstvty ndex of node for the total response varaton of the system could be normalzed by usng Eq. (3). The normalzed senstvty ndex of node s expressed n percentage and represents the senstvty proporton of a node pont n the system. Table 1 shows the results of the normalzed senstvty analyss of the lower control arm and the subframe at 225 Hz on the tests. The results were expressed n descendng order wth the node numbers and normalzed senstvty values of both experments. Fgure 6 shows that the lower control arm has a large deformaton around node number 1 on both tests. The normalzed senstvty analyss results smlarly showed that the part around node number 1 had the largest senstvty n both experments. The ODS result s consequently verfed by these desgn senstvty analyss results. The ODS and the normalzed senstvty n the sub-frame also almost had the same results n both tests. As mentoned earler, the sub-frame at 225 Hz had a torsonal dynamc moton, and the left and rght front mount ponts moved the most. The desgn senstvty analyss results had the largest value at node numbers 11 and 10 located on the rght mount ponts. Ths fndng verfed that the ODS and desgn senstvty analyss results had the same tendency. Furthermore, the rg test could represent the same dynamc moton as that n the baselne vehcle test. Table 1 Normalzed senstvty analyss results of the left lower control arm and the sub-frame on the baselne vehcle and the rg at 225 Hz 225 Hz Lower control arm (left) Sub-frame Order Car Rg Car Rg Node # Node # Node # Node # 1 # # # # # # # # # # # # # # # # # # # # CONCLUSIONS The expermental method used to analyze the structure-borne road nose of a suspenson was proposed by usng a specal suspenson rg. Ths suspenson rg made t possble to analyze the road nose performance of a vehcle suspenson, excludng the effect of the vehcle body, and structurally evaluate the suspenson component factors affectng the road nose through ODS measurement. Ths method was valdated by comparng the ODS measured n both the baselne vehcle and suspenson rg tests. 5342

7 The ODS comparson results showed that the frequences of the major contrbuton peas, such as around 8 and 22, whch are tre cavty-resonant, were the same although the dynamc moton magntude of the rg was bgger than that of the baselne vehcle. The dynamc motons of the suspenson sub-frame and the lower control arms at the peas were also the same n both cases based on the senstvty analyss results. In concluson, ths research could help n performng an effcent road nose analyss of a suspenson by combnng the evaluaton method of the structural dynamc moton wth the TPA analyss usng the suspenson rg proposed by Davd [1]. Ths research could also help reduce the road nose evaluaton tme by completng the development stage pror to the body prototype development n vehcle NVH-related felds. ACKNOWLEDGEMENTS Ths research was supported by the NVH research laboratory n Hyunda Motor Company. REFERENCES 1. Song, Davd P., et al. "A methodoloy for evaluatng structure-borne road nose pror to a development mule usng drect force measured on a suspenson rg."inter-noise and NOISE-CON Congress and Conference Proceedngs. Vol No. 5. Insttute of Nose Control Engneerng, Km, Chan-Jung, et al. "Desgn senstvty analyss of a system under ntact condtons usng measured response data." Journal of Sound and Vbraton (2012): Kromuls, J., and E. Hojan. "An applcaton of two expermental modal analyss methods for the determnaton of operatonal deflecton shapes."journal of Sound and Vbraton (1996):