外文翻譯---液壓密封完整性調(diào)查研究

1、<p>　　Mechanical Systems and Signal Processing, 2007, 21: 1115–1126</p><p>　　A study of hydraulic seal integrity</p><p>　　P. Chena, P.S.K. Chua, G.H. Lim</p><p>　　Abstract: The

2、 work described in this paper involved on-line detection of seal defects in a water hydraulic cylinder. An obvious effect of seal defect is internal leakage. Therefore, the approach used was to detect the internal leakag

3、e using suitable technique. The technique used involved detecting the acoustic emission (AE) due to the internal leakage. This paper evaluated various parameters of AE signals in terms of their capability in estimating t

4、he internal leakage rate in a water hydraulic c</p><p>　　Keywords: Acoustic emission; Water hydraulic cylinder; Internal leakage; AE count rate; Root mean square; Power spectral density; AE energy</p>

5、<p>　　1. Introduction</p><p>　　Modern water hydraulics, using tap water as the hydraulic fluid, has gained much interest in the past decade due to its inherent advantages compared to oil hydraulics. Th

6、ese advantages include environment friendliness, good product compatibility and no fire hazards [1, 2]. However, some problems with modern water hydraulics are still to be addressed. One of the most common problems is th

7、e relatively large internal leakage in water hydraulic components. For example, a water hydraulic cylinder cou</p><p>　　The work presented in this paper is part of a project that aims to develop a quantitati

8、ve model to estimate the internal leakage flow rate in a water hydraulic cylinder by means of AE. It is focused on the internal leakage smaller than 1.0 L/min. In order to model the AE signal generated by the internal le

9、akage, suitable parameters must first be selected to interpret the signal. Therefore, experiments were conducted to study the characteristics of various AE parameters in terms of their effectiv</p><p>　　2. A

10、coustic emission</p><p>　　AE is defined as the transient elastic waves that are generated by the rapid release of energy from localised sources. It has been found that AE signals can be generated by fluid le

11、akage. Pollock and Hsu [10] studied the physical origin of these signals in detail and Goodman et al. [12] reported a variety of AE source mechanisms associated with leakage from vessels, tanks and pipelines. In the case

12、 of internal leakage in water hydraulic cylinders, the generation of AE signals is largely attribut</p><p>　　AE counts are widely used as a practical measure of AE activity. This parameter is defined as the

13、number of times the signal exceeds a counter threshold. For continuous-type AE, AE count rate is often used to measure the variation of AE counts with time. The root-mean-square (rms) value is often used to measure the e

14、nergy content of AE signals. For an AE signal consisting ofx [0]， x [1], ……, x [N?1] , its rms value is</p><p>　　The advantage of energy measurement is that the energy content of the AE signal can be directl

15、y related to important physical parameters associated with the energy release at the AE source [14]. The above parameters have been used to describe AE signals in a variety of applications [11, 17, 18].</p><p&

16、gt;　　The aforementioned parameters are measured in the time domain. Besides, parameters measured in the frequency domain are also of interest, such as the frequency and magnitude of the dominant frequency component and t

17、he energy contained within frequency bands. For the continuous-type AE, these parameters can be obtained through spectral analysis using Fourier transform. The power spectral density (PSD) of AE signals can be computed u

18、sing the following equation [19]:</p><p>　　where P[ k ] is the power spectral density, X [k]] is the discrete Fourier transform (DFT) of an AE signal x[n], andT is the sampling period. The PSD represents the

19、 distribution of the signal power over frequencies. Some studies of AE signals in the frequency domain can be found in Refs. [10, 13, 20, 21].</p><p>　　3. Experimentation</p><p>　　Due to the com

20、plexity of AE phenomena, analytical methods are not well established. Therefore, experimental methods are introduced to investigate AE. In order to study the characteristics of AE signals generated by internal leakage in

21、 water hydraulic cylinders, experiments were deliberately designed, as described below.</p><p>　　For each record of AE signal, the AE count rate, denoted as _N AE was calculated by dividing the AE counts by

22、the signal duration. Both a fixed threshold and a floating threshold were used for counting. Since there was no well-defined procedure to choose the threshold value, a wide range of values were tried. For the fixed thres

23、hold, a value of 0.04V yielded the best results, as shown in Fig. 6a. It is noted that the AE count rate drops fast as the internal leakage rate decreases. For the floati</p><p>　　In order to simulate scores

24、 created by the abrasive action of solid particulates, a file was used, in the present work, to make scores on the piston seal surfaces of a water hydraulic cylinder. Fig. 2 shows the scored piston seals used in the expe

25、riments. These seals lead to an internal leakage smaller than 1.0 L/min for the pressure range of 0–70 bar. Sixteen scores were equally distributed along the circumference of the seals. The dimensions of these scores wer

26、e measured with a non-contact opt</p><p>　　Fig. 2. The 16-score piston seals.</p><p>　　Fig. 3. The profile of a score.</p><p>　　4. Experimental results</p><p>　　In the

27、experiment, 100 sets of data were acquired at different internal leakage rates, with each set consisting of 40 records of AE signals measured at a certain leakage rate. Each record of AE signal contained 4096 points samp

28、led at 5 MHz, from which AE parameters were calculated. For each AE parameter, results obtained from the 40 records were then averaged. In the following, all the results are the average values.</p><p>　　For

29、each record of AE signal, the AE count rate, denoted as _N AE, was calculated by dividing the AE counts by the signal duration. Both a fixed threshold and a floating threshold were used for counting. Since there was no w

30、ell-defined procedure to choose the threshold value, a wide range of values were tried. For the fixed threshold, a value of 0.04V yielded the best results, as shown in Fig. 4a. It is noted that the AE count rate drops fa

31、st as the internal leakage rate decreases. </p><p>　　Fig. 4. AE count rate versus internal leakage rate</p><p>　　For the floating threshold, the threshold value was set to be proportional to the

32、 rms value of the signal. The resulting AE count rate remained at a constant level, nom atter how the leakage rate varied. This is shown in Fig. 4b, where the AE count rate was obtained with the threshold equal to the rm

33、s value of the signal. It can be seen that there is no desirable trend in the AE count rate with respect to the leakage rate.</p><p>　　5. Predict the internal leakage rate</p><p>　　As has been s

34、hown in the above, the energy content of AE signal is closely related to the internal leakage rate in the water hydraulic cylinder. Therefore, it may be used to predict the internal leakage rate. The error of prediction,

35、 then, is of interest. In the following, an empirical model is built to predict the internal leakage rate based on measured AE signals and the error of prediction is analysed with statistical methods. Due to the simplici

36、ty in calculation, the rms value Vrms is chosen</p><p>　　Qi=7.86Vrms+0.14. </p><p>　　For a measured AE rms value, the internal leakage rate may be predicted with Eq.</p><p>　　(7). S

37、uppose the measured AE rms value is Vrms0. A 95% prediction interval for the true value of the internal leakage rate,denoted as Qi0, is given by</p><p>　　where ^Qi is the internal leakage rate predicted by E

38、q. (7) based on the measured Vrms0 and d is a measure of the width of the prediction interval. Note that d is not a constant but varies with the measured AE rms value Vrms0. For the range of the internal leakage rates sm

39、aller than 1.0 L/min, d is about 0.078 L/min. Eq. (8) means that for the measured AE rms value Vrms0, the true value of the internal leakage rate Qi0 lies inside the intervale ^Qi d; ^Qi t dT with 95% confidence. </p

40、><p>　　6. Conclusions</p><p>　　This paper analysed the characteristics of AE signals generated by internal leakage in a water hydraulic cylinder. Experiments were carefully designed, including the

41、simulation of the internal leakage across the piston seals in a water hydraulic cylinder and the measurement of the internal leakage rate. AE signals obtained from the experiments were analysed, in which several AE param

42、eters were extracted from the AE signals and their effectiveness for predicting the internal leakage rate were stu</p><p>　　(1) AE signals are sensitive to small internal leakage in a water hydraulic cylinde

43、r and AE-based methods are able to predict the internal leakage that is smaller than 1.0 L/min.</p><p>　　Energy-based AE parameters, whether measured in the time domain or in the frequency domain, are more s

44、uitable than the AE count rate and the peak PSD magnitude to interpret AE signals generated by the internal leakage.</p><p>　　References</p><p>　　[1] G.W. Krutz, P.S.K. Chua, Water hydraulics—th

45、eory and applications 2004, in: Proceedings of the Workshop on Water Hydraulics, Agricultural Equipment Technology Conference (AETC ’04), Louisville, KY, USA, February 8–10, 2004.</p><p>　　[2] E. Trostmann,

46、Water Hydraulics Control Technology, Marcel Dekker, New York, USA, 1996.</p><p>　　[3] W. Backe′ , Water- or oil-hydraulics in the future, in: Proceedings of the Sixth Scandinavian International Conference on

47、 Fluid Power, Tampere, Finland, May 26–28, 1999, pp. 51–64.</p><p>　　[4] J. Watton, Condition Monitoring and Fault Diagnosis in Fluid Power Systems, Ellis Horwood, New York, USA, 1992.</p><p>　　

48、[5] T.T. Le, J. Watton, D.T. Pham, An artificial neural network based approach to fault diagnosis and classification of fluid power systems, Proceedings of the Institution of Mechanical Engineers, Part I, Journal of Syst

49、ems and Control Engineering 211 (1997)</p><p><b>　　307–317.</b></p><p>　　[6] T.T. Le, J. Watton, D.T. Pham, Fault classification of fluid power system using a dynamics feature extrac

50、tion technique and neural networks, Proceedings of the Institution of Mechanical Engineers, Part I, Journal of Systems and Control Engineering 212 (1998) 87–97.</p><p>　　[7] G. Thompson, G. Zolkiewski, An ex

51、perimental investigation into the detection of internal leakage of gases through valves by vibration analysis, Proceedings of the Institution of Mechanical Engineers, Part E, Journal of Process Mechanical Engineering 211

52、 (1997) 195–207.</p><p>　　[8] M. Pietola, R. Ma¨ kinen, P. Va¨ yrynen, S. Kesanto, J. Varrio, Using a high resolution thermograph in predictive maintenance and fault diagnosis of fluid power compon

53、ents and systems, in: Proceedings of the Fourth Scandinavian International Conference on Fluid Power, Tampere, Finland, September 26–29, 1995, pp. 719–725.</p><p>　　機械系統(tǒng)與信號處理, 2007, 21: 1115–1126</p>

54、<p>　　液壓密封完整性調(diào)查研究</p><p>　　P. Chena, P.S.K. Chua, G.H. Lim</p><p>　　摘要：本文中所涉及在液壓缸的上線檢測密封缺陷. 一個明顯的影響密封的缺陷是內(nèi)部泄漏。因此，所采用的辦法是使用合適的技術(shù)探測內(nèi)部泄漏. 所采用的技術(shù)涉及由于檢測聲發(fā)射（ AE ）內(nèi)部泄漏.本文評估了AE信號的各種各樣的參量,根據(jù)他們估計液壓缸內(nèi)

55、部漏出率。 實驗分析了AE參量不同的內(nèi)部漏出率，參量包括根均方(rms)值，計數(shù)率、繁忙程度功率譜密度和能量的特征。分析了這些參量和內(nèi)部漏出率之間的交互關(guān)系。 結(jié)果表示，基于能量的AE參量，特別是均方根值，是更加適當解釋內(nèi)部漏出引起的AE信號。</p><p>　　關(guān)鍵詞：聲發(fā)射；液壓缸；內(nèi)部泄漏；聲發(fā)射計數(shù)率；均方根功率譜密度；AE能量</p><p><b>　　1 引言&l

56、t;/b></p><p>　　現(xiàn)代水利使用自來水作為液壓油，在過去幾十年中，由于液壓油其固有的優(yōu)勢相比。這些優(yōu)勢包括環(huán)境，友善，良好的產(chǎn)品兼容性，并沒有發(fā)生火警的危險。但是，一些問題，與現(xiàn)代水力仍有待解決。其中最常見的問題是比較大的內(nèi)部滲漏水液壓元件。舉例來說，一個水上液壓缸可能遭受橫跨活塞封印的內(nèi)部漏出。這歸結(jié)于非常低粘度的水力液壓油。因此，重要的是要監(jiān)測內(nèi)部漏出以達到最佳性能和水液壓機構(gòu)的可靠和安全。

57、</p><p>　　本文提出的是打算開發(fā)一個定量模型通過AE估計在水液壓缸的內(nèi)部漏出流速項目的一部分。它集中于內(nèi)部漏出小于1.0 L /min。為了塑造AE發(fā)信號引起由內(nèi)部漏出，適當?shù)膮⒘勘仨毷紫冗x擇解釋信號。所以，試驗根據(jù)他們的在估計內(nèi)部漏出率的有效率做了各種各樣的AE參量的特征，正如本文所描述。</p><p><b>　　2. 聲發(fā)射</b></p>

58、;<p>　　AE被定義作為由迅速能量的釋放從局部性引起的瞬變彈性波。AE信號產(chǎn)生液體泄漏?？?。波洛克和古德 曼等人詳細研究了這些信號的物理起源，并且Goodman等報告了各種各樣的AE來源機制與從船、坦克和管道的漏出相關(guān)。 在液壓缸的內(nèi)部漏出情況下，AE信號主要歸因于內(nèi)部漏出導(dǎo)致的動蕩。</p><p>　　聲發(fā)射信號可分為兩種基本類型。爆裂型聲發(fā)射是指聲發(fā)射信號所對應(yīng)的個人聲發(fā)射事件，而連續(xù)型聲

59、發(fā)射指的顯然是持續(xù)的信號水平迅速發(fā)生聲發(fā)射活動。聲發(fā)射信號的產(chǎn)生是由內(nèi)部滲漏水液壓缸連續(xù)型，如圖1所示。</p><p>　　圖1：AE信號所產(chǎn)生的內(nèi)部滲漏水液壓缸</p><p>　　一般來說，直接觀察的連續(xù)聲發(fā)射信號很少關(guān)于AE信號源資料。為了提取更多有用的信息，由聲發(fā)射信號，首先應(yīng)該適當?shù)亟忉屝盘枺ǔＩ婕懊枥L他們與有些參量。為定量AE調(diào)查，任何一個數(shù)學(xué)模型進行之前參數(shù)必須加以界定，

60、。各項參數(shù)已用于AE信號特性，無 論是在時域和頻域。以下簡單地描述有些用途廣泛的參量連續(xù)式AE發(fā)信號。AE計數(shù)用途廣泛，是作為AE活動一項實用措施。 當次數(shù)信號超出逆門限，這個參量被定義。 對于連續(xù)式AE， AE計數(shù)率是常用的測量AE計數(shù)的變異與時間的。均方根(rms)是常用的測量AE信號能量內(nèi)含。 對于包括N樣品的AE信號，它的rms值 x [0]， x [1]，……， x [N?1]其有效值為</p><p>

61、;<b>　　（1）</b></p><p>　　能量測量的優(yōu)點是AE信號的能量內(nèi)含可以直接地與重要物理參量釋放能量。上述參量被廣泛應(yīng)用于描述AE信號。上述的參量在時間界域被測量。 其外，在頻域測量的參量也是利益，例如在頻帶內(nèi)和巨大包含的統(tǒng)治頻率組分和能量的頻率。 使用傅立葉變換，對于連續(xù)式AE，這些參量可以通過光譜分析得到。功率譜密度（ PSD ）的聲發(fā)射信號，可使用以下公式連續(xù)計算<

62、;/p><p><b>　　(2)</b></p><p>　　其中P[ k ]是功率譜密度，X [k]是分離傅立葉變換(DFT) AE信號x [n]，T是取樣周期。PSD的代表分布的信號功率超過頻率。AE信號的有些研究在頻域的可以在Refs找到。</p><p><b>　　3. 實驗</b></p><

63、p>　　由于AE現(xiàn)象的復(fù)雜，沒有固定得分析方法。 所以，介紹實驗法調(diào)查AE。 為了學(xué)習(xí)AE信號的特征在液壓缸的內(nèi)部漏出引起的，實驗設(shè)計，如下所述：</p><p>　　在液壓缸的內(nèi)部漏出通過連接流量控制閥模仿了與圓筒平行。人工介紹的內(nèi)部漏出流經(jīng)了閥門而不是液壓缸。 這種的優(yōu)點是模仿漏出率可能容易地是受控的。 然而，模仿的有效性要求進一步調(diào)查。相信這種方法不可能提供存在于圓筒的動態(tài)過程的中意的模仿受內(nèi)部漏出支

64、配。 因此，在當前工作，努力被做了模仿在液壓缸的真正的內(nèi)部漏出。 首先研究下面漏出機制然后，然后提出漏出的模仿。</p><p>　　為了模仿堅實微粒物質(zhì)的磨蝕行動創(chuàng)造的比分，文件在當前工作在活塞被用于，做密封水液壓缸的表面。 圖2顯示用于實驗的被計分的活塞封印。 這些封印帶領(lǐng)內(nèi)部漏出0–70酒吧的壓力范圍的小于1.0升/分鐘。16個得分沿封印的圓周平等地被分布了。 維度這些比分測量了與一個沒有接觸的光學(xué)測量系統(tǒng)

65、。 圖3顯示測量系統(tǒng)采取的比分的外形。 沿比分的邊緣，五個關(guān)鍵被選擇了，并且測量了他們的座標。 然后測量了得分的寬度和深度。另外，計算了這五點的圓弧適合。 因此，比分的一條近似半徑能獲得。</p><p>　　圖2 16比分活塞封印</p><p>　　圖3 比分的外形</p><p><b>　　4 . 實驗結(jié)果</b></p&

66、gt;<p>　　在實驗中，100套數(shù)據(jù)獲取不同的內(nèi)部泄漏率，當每個集合包括AE信號40個紀錄被測量以某一漏出率。每條記錄的AE信號載4096點采樣5兆赫，從哪個聲發(fā)射參數(shù)計算。每個聲發(fā)射參數(shù)，從40個紀錄得到的結(jié)果然后平均，下面所有結(jié)果是平均值。</p><p>　　AE信號的每條記錄，AE計數(shù)率，簡稱能量.通過劃分由信號期間的AE計數(shù)計算。使用了固定的門限和浮動門限為計數(shù)。因為沒有選擇閾值的明確

67、定義的做法，各種各樣的價值被嘗試了。如圖4a所顯示，為固定閾值，0.04 V的價值產(chǎn)生了最佳的結(jié)果。 注意到，聲發(fā)射計數(shù)率的下降速度是內(nèi)部泄漏率降低。 對于浮動門限，設(shè)置閾值是比例與信號的rms價值。由此產(chǎn)生的AE計數(shù)率保持在一個恒定的水平，不管漏出率變化。如圖4b顯示，AE計數(shù)率獲得與閾值相等與信號的rms價值。能看見沒有關(guān)于漏出率在AE計數(shù)率的中的趨向。</p><p>　　圖4. AE計數(shù)率對內(nèi)部漏出率&l

68、t;/p><p><b>　　5 預(yù)測內(nèi)部泄漏率</b></p><p>　　上述情況表明，能源含量的AE信號是于液壓缸的內(nèi)部泄漏率密切相關(guān)的。因此，它可以被用來預(yù)測內(nèi)部泄漏率。在隨后的一個模型是建立在基于實測聲發(fā)射信號與預(yù)測誤差分析與統(tǒng)計方法預(yù)測的內(nèi)部泄漏率。由于要求計算簡單，是選擇均方根值而不是特征聲發(fā)射信號。從以前的實驗數(shù)據(jù)之間的關(guān)系，利用最小二乘法給出</p

69、><p>　　Qi=7.86Vrms+0.14. （3）</p><p>　　為被測量的AE 均方根值，內(nèi)部漏出率預(yù)言Eq。 假設(shè)被測量的AE均方根值是Vrms0。 表示作為Qi0，測量內(nèi)部漏出率的真實值的95%預(yù)言間隔時間</p><p><b>　?。?a）</b></p&g

70、t;<p><b>　　(4b) </b></p><p>　　那里是Eq預(yù)測的內(nèi)部漏出率。 根據(jù)測量的Vrms0和δ預(yù)計間隔時間的寬度的措施。 注意δ不是常數(shù)，但是隨被測量的AE均方根值Vrms0變化。內(nèi)部漏出的范圍估計小于1.0升/分鐘， δ大約0.078升/每分鐘。 (4a)和(4b)意味那被測量的AE rms價值的Vrms0，內(nèi)部漏出率Qi0的真實值在間隔時間里面的有9

71、5%的可能。 關(guān)于預(yù)言間隔時間的演算的細節(jié)可以在參考找到。圖11顯示內(nèi)部漏出率的間隔時間。 δ的值反映在估計內(nèi)部漏出率的錯誤。 實際上，0.078 升/分鐘在液壓機構(gòu)的內(nèi)部漏出是可以接受的。經(jīng)驗?zāi)Ｐ徒o出由Eq是重要的。 不可以適用其他情況。 事實上，由于這個相當復(fù)雜的聲發(fā)射現(xiàn)象問題，一般的聲發(fā)射模式尚未確定，即適用于任何情況。因此，應(yīng)用聲發(fā)射技術(shù)是顯著依賴于具體的情況。</p><p><b>　　6.

72、 結(jié)論</b></p><p>　　本文分析了液壓缸的內(nèi)部漏出是AE信號的特征引起的。試驗是經(jīng)過精心設(shè)計，包括液壓缸內(nèi)部漏出和模仿活塞密封內(nèi)部漏出率的測量。 分析了從實驗得到的AE信號，幾個參量從AE信號提取，并且研究了他們預(yù)測的內(nèi)部漏出率。從分析結(jié)果，可以得到如下結(jié)論：</p><p>　　基于能量的AE參量，測量在時間域或在頻域，是否比AE計數(shù)率和峰頂PSD解釋內(nèi)部漏出適當

73、引起的AE信號。</p><p>　　AE rms價值有與內(nèi)部漏出率的牢固的線性關(guān)系在水液壓缸，并且，它是的一個最中意的參量在一個定量AE模型的發(fā)展確定內(nèi)部漏出率。</p><p><b>　　參考文獻</b></p><p>　　[1] g.w. krutz, 蔡銳明. 水力學(xué)的理論與應(yīng)用, 2004, 路易斯維爾, KY, 美國, 2月8日

74、至10日, 2004.</p><p>　　[2] E. Trostmann. 水動水學(xué)控制技術(shù). 美國紐約,1996.</p><p>　　[3] J. watton. 狀態(tài)監(jiān)測和故障診斷中的流體動力系統(tǒng). 埃利斯horwood，New York，美國, 1992年.</p><p>　　[4]亨特. 泄漏及流量監(jiān)測. 英國牛津, 2001年.</p>

75、<p>　　[5]波洛克，許信良. 檢漏用聲發(fā)射，1982.</p><p>　　[6] A.A.狹鱈. 從氣體泄漏的聲發(fā)射的定量分析. 第18卷, Kluwer院出版者, 美國紐約(1999) .</p><p>　　[7]癥古德曼, k.米勒, 查韋斯科爾. 聲學(xué)泄漏檢測, 美國, 1998年, 第11章.</p><p>　　[8]德魯亞爾. 歷