The relationship between echolocation frequency (represented by dominant frequency, DF for short) and body size (body mass, forearm length and body length) in 8 species of horseshoe bats ( Rhinolophus cornutus, R. aff...The relationship between echolocation frequency (represented by dominant frequency, DF for short) and body size (body mass, forearm length and body length) in 8 species of horseshoe bats ( Rhinolophus cornutus, R. affinis, R. thomasi, R. rouxi, R. blythii, R. ferrumequinum, R. pearsoni, R. rex ) was examined. The eight species were captured in caves in five regions (Guiyang, Anlong, Xingyi, Anshun and Zhenning) of Guizhou Province in July and August 1999 and June 2000. The species were identified based on the descriptions in Mammals of GuiZhou (Luo et al .)and Key to the Identification of Chiroptera (Wang, unpublished). The bats were captured at the entrance to the caves at about 8 pm (the time when bats usually leave the caves), and were then put into a recording room near the capture locations where they could fly freely. Recordings of their echolocation calls were made bout 2 hours later using an ultra sound detector (U30, Ultra Sound Advice, UK) which recorded the calls of bats that were flying directly at the microphone at a distance of 1 m. Each bat was recorded 5 times and the signals were fed into a portable ultra sound processor (PUSP, Ultra Sound Advice, UK). The duration for recording was 1 1 s with a sampling frequency of 44 1 kHz. By replaying the recorded tapes the best quality recordings (the loudest and clearest with the least background noise) were replayed at 1/10 speed and re recorded using a digital sound recorder (Sony, MD 1, frequency response range: 30～20 000 Hz). The re recorded echolocation signals were analyzed using the sound processing software Cool Edit 2000, developed by the American Syntrillium Software Company. Ultra sound analysis referred to the sound spectrograms (frequency time graph), time domain spectrograms (energy time graph), energy spectrograms (energy frequency graph), and Hanning window to obtain an analytic precision of 256 Hz. The analysis attenuation was 60 dB. The DF, pulse duration and interpulse interval of the echolocation calls were recorded and the duty cycle, which represents the percentage of the pulse duration in the summation of the pulse duration and the interpulse interval, was calculated. The data are presented as +SD . Body size were measured using a vernier caliper and a balance; all measurements are presented as +SD. When flying, all eight species of bats had similar echolocation signals: the pattern of their echolocation calls was (FM ) CF FM (FM means Frequency modulated; CF means Constant frequency) with 1～2 harmonics. Pulse duration was more than 10 ms, the duty cycle was higher than 40%, the dominant frequency mainly concentrated on the CF part from 25 kHz to 120 kHz. The different species did, however, display different FM widths. Pearson (one of analytical methods in the software SPSS 10 0) was adopted to analyze the correlation between dominant frequency and body size. An obvious negative correlation was found between echolocation frequency and body size in horseshoe bats. The correlation coefficient of DF against forearm length was r=-0 714 (P=0 047; df=6) , DF against body mass r= -0 429 (P=0 289; df=6) and DF against body length r=-0 810 (P=0 015, df=6) . The dominant frequency was higher in smaller species. The species order in terms of dominant frequencies (highest to lowest) was: Rhinolophus cornutus>R. thomasi>R. rouxi>R. blythi>R.ferrumequinum>R. pearsoni>R. affinis>R. rex. The order of forearm length from top to bottom was: Rhinolophus pearsomi>R. rex>R. ferrumequinum>R. affinis>R. thomasi>R. rouxi>R. cornutus>R. blythi. The negative correlation between the dominant frequency and body size was not very strong, presumably a reflection of the influence of factors such as ecological competition and morphological and physiological features other than body size on echolocation calls. Divergence in echolocation calls allows bat species of similar body size to avoid competition with each other. The relationship between dominant frequency and body size probably occurs because the wave length of展开更多
The studies on dynamics of a fault bearing system are prevalent in recent years, however, we are studying a completely different frequency range than the one where the bearing faults are best seen. Considering a local...The studies on dynamics of a fault bearing system are prevalent in recent years, however, we are studying a completely different frequency range than the one where the bearing faults are best seen. Considering a local defect on outer raceway,a two-degree-of-freedom analytical model of a rigid-rotor ball bearing system is established. Three pulse force models are introduced to simulate the local defect. The frequency domain method—harmonic balance method with alternating frequency/time domain technique (HB-AFT) is used to calculate the response in a large frequency range. By comparing the performance at different frequencies, the fault systems with different defect models and parameters reveal the super-harmonic resonances,and the reasons for this phenomenon are uncovered as well. Finally, the theoretical calculation is verified qualitatively by the experimental results, through comparing the frequency spectrums of the defective bearing rotor system to the fault-free one.Therefore, the super-harmonic resonances can be regarded as a dynamic feature. Besides, the obvious super-harmonic resonances indicate the magnification of the harmonics of the "characteristic defect frequency" for outer race in the corresponding speed regions, which may be helpful for the diagnosis of a rotor ball bearing system with a local defect.展开更多
文摘The relationship between echolocation frequency (represented by dominant frequency, DF for short) and body size (body mass, forearm length and body length) in 8 species of horseshoe bats ( Rhinolophus cornutus, R. affinis, R. thomasi, R. rouxi, R. blythii, R. ferrumequinum, R. pearsoni, R. rex ) was examined. The eight species were captured in caves in five regions (Guiyang, Anlong, Xingyi, Anshun and Zhenning) of Guizhou Province in July and August 1999 and June 2000. The species were identified based on the descriptions in Mammals of GuiZhou (Luo et al .)and Key to the Identification of Chiroptera (Wang, unpublished). The bats were captured at the entrance to the caves at about 8 pm (the time when bats usually leave the caves), and were then put into a recording room near the capture locations where they could fly freely. Recordings of their echolocation calls were made bout 2 hours later using an ultra sound detector (U30, Ultra Sound Advice, UK) which recorded the calls of bats that were flying directly at the microphone at a distance of 1 m. Each bat was recorded 5 times and the signals were fed into a portable ultra sound processor (PUSP, Ultra Sound Advice, UK). The duration for recording was 1 1 s with a sampling frequency of 44 1 kHz. By replaying the recorded tapes the best quality recordings (the loudest and clearest with the least background noise) were replayed at 1/10 speed and re recorded using a digital sound recorder (Sony, MD 1, frequency response range: 30～20 000 Hz). The re recorded echolocation signals were analyzed using the sound processing software Cool Edit 2000, developed by the American Syntrillium Software Company. Ultra sound analysis referred to the sound spectrograms (frequency time graph), time domain spectrograms (energy time graph), energy spectrograms (energy frequency graph), and Hanning window to obtain an analytic precision of 256 Hz. The analysis attenuation was 60 dB. The DF, pulse duration and interpulse interval of the echolocation calls were recorded and the duty cycle, which represents the percentage of the pulse duration in the summation of the pulse duration and the interpulse interval, was calculated. The data are presented as +SD . Body size were measured using a vernier caliper and a balance; all measurements are presented as +SD. When flying, all eight species of bats had similar echolocation signals: the pattern of their echolocation calls was (FM ) CF FM (FM means Frequency modulated; CF means Constant frequency) with 1～2 harmonics. Pulse duration was more than 10 ms, the duty cycle was higher than 40%, the dominant frequency mainly concentrated on the CF part from 25 kHz to 120 kHz. The different species did, however, display different FM widths. Pearson (one of analytical methods in the software SPSS 10 0) was adopted to analyze the correlation between dominant frequency and body size. An obvious negative correlation was found between echolocation frequency and body size in horseshoe bats. The correlation coefficient of DF against forearm length was r=-0 714 (P=0 047; df=6) , DF against body mass r= -0 429 (P=0 289; df=6) and DF against body length r=-0 810 (P=0 015, df=6) . The dominant frequency was higher in smaller species. The species order in terms of dominant frequencies (highest to lowest) was: Rhinolophus cornutus>R. thomasi>R. rouxi>R. blythi>R.ferrumequinum>R. pearsoni>R. affinis>R. rex. The order of forearm length from top to bottom was: Rhinolophus pearsomi>R. rex>R. ferrumequinum>R. affinis>R. thomasi>R. rouxi>R. cornutus>R. blythi. The negative correlation between the dominant frequency and body size was not very strong, presumably a reflection of the influence of factors such as ecological competition and morphological and physiological features other than body size on echolocation calls. Divergence in echolocation calls allows bat species of similar body size to avoid competition with each other. The relationship between dominant frequency and body size probably occurs because the wave length of
基金supported by the National Key Basic Research Program of China (Grant No. 2015CB057400)the National Natural Science Foundation of China (Grant No. 11602070)China Postdoctoral Science Foundation(Grant No. 2016M590277).
文摘The studies on dynamics of a fault bearing system are prevalent in recent years, however, we are studying a completely different frequency range than the one where the bearing faults are best seen. Considering a local defect on outer raceway,a two-degree-of-freedom analytical model of a rigid-rotor ball bearing system is established. Three pulse force models are introduced to simulate the local defect. The frequency domain method—harmonic balance method with alternating frequency/time domain technique (HB-AFT) is used to calculate the response in a large frequency range. By comparing the performance at different frequencies, the fault systems with different defect models and parameters reveal the super-harmonic resonances,and the reasons for this phenomenon are uncovered as well. Finally, the theoretical calculation is verified qualitatively by the experimental results, through comparing the frequency spectrums of the defective bearing rotor system to the fault-free one.Therefore, the super-harmonic resonances can be regarded as a dynamic feature. Besides, the obvious super-harmonic resonances indicate the magnification of the harmonics of the "characteristic defect frequency" for outer race in the corresponding speed regions, which may be helpful for the diagnosis of a rotor ball bearing system with a local defect.
