水泵水輪機(jī)甩負(fù)荷過(guò)程流動(dòng)誘導(dǎo)噪聲數(shù)值模擬

毛秀麗，孫奧冉，Giorgio Pavesi，鄭 源，葛新峰

?

水泵水輪機(jī)甩負(fù)荷過(guò)程流動(dòng)誘導(dǎo)噪聲數(shù)值模擬

毛秀麗1，孫奧冉2，Giorgio Pavesi3，鄭 源4，葛新峰4

（1. 西北農(nóng)林科技大學(xué)水利與建筑學(xué)院，楊凌 712100；2.安徽省水利水電勘測(cè)設(shè)計(jì)院，合肥 230088；3. 帕多瓦大學(xué)工業(yè)工程學(xué)院，帕多瓦 35131；4. 河海大學(xué)水利水電學(xué)院，南京 210000）

為研究水泵水輪機(jī)甩負(fù)荷過(guò)程壓力脈動(dòng)特性及其流動(dòng)誘導(dǎo)噪聲，該文基于網(wǎng)格壁面滑行技術(shù)和DES湍流模型，對(duì)水泵水輪機(jī)發(fā)電工況下導(dǎo)葉關(guān)閉過(guò)程進(jìn)行連續(xù)性模擬，并將流場(chǎng)葉片表面壓力脈動(dòng)信號(hào)作為聲場(chǎng)流動(dòng)誘導(dǎo)噪聲計(jì)算聲源，通過(guò)對(duì)壓力脈動(dòng)特性和流動(dòng)誘導(dǎo)噪聲分析得到：導(dǎo)葉進(jìn)出口處2個(gè)無(wú)葉區(qū)內(nèi)壓力脈動(dòng)主頻位置均在葉頻為斯特勞哈爾數(shù)等于0.051與1處，導(dǎo)葉出口處頻譜值是進(jìn)口處10倍之上，說(shuō)明動(dòng)動(dòng)干涉對(duì)流態(tài)的影響強(qiáng)于動(dòng)靜干涉；當(dāng)尾水管內(nèi)出現(xiàn)2個(gè)反向旋壁渦帶時(shí)，壓力脈動(dòng)最強(qiáng)烈且其幅值最大。聲場(chǎng)分析結(jié)果表明外場(chǎng)噪聲的主頻由壓力脈動(dòng)主頻與殼體固有頻率綜合決定，聲壓分布具有“∞”形式的指向形態(tài)，且各階葉頻處聲壓分布呈現(xiàn)出明顯的對(duì)稱性，說(shuō)明葉片噪聲輻射具有明顯的偶極子特性；在一階、二階葉頻處，導(dǎo)葉關(guān)閉前一半階段，流量對(duì)于外場(chǎng)噪聲輻射能力的影響表現(xiàn)為大流量工況下最強(qiáng)，小流量工況下最弱，導(dǎo)葉關(guān)閉后一半階段正好相反。

壓力；葉輪；噪聲；水泵水輪機(jī)；甩負(fù)荷過(guò)程

0 引 言

抽水蓄能電站機(jī)組運(yùn)行穩(wěn)定性一直是當(dāng)下業(yè)內(nèi)研究的熱點(diǎn)和難點(diǎn)所在，而機(jī)組過(guò)渡過(guò)程是影響其穩(wěn)定運(yùn)行的一個(gè)重要因素[1]。尤其近年來(lái)新能源的蓬勃發(fā)展，抽蓄機(jī)組需要更頻繁的在水泵和水輪機(jī)工況之間切換以滿足電網(wǎng)的需求[2]。水泵水輪機(jī)在過(guò)渡過(guò)程中運(yùn)行時(shí)，內(nèi)部復(fù)雜流態(tài)可量化表現(xiàn)為壓力脈動(dòng)現(xiàn)象，用以反映機(jī)組穩(wěn)定性[3-4]，因而對(duì)水泵水輪機(jī)過(guò)渡過(guò)程計(jì)算研究意義重大。針對(duì)上述問(wèn)題已有一定研究[5-7]，然而前人的工作主要從流場(chǎng)角度分析[8-9]，對(duì)于機(jī)組穩(wěn)定運(yùn)行所開(kāi)展的工作較少，為改善抽蓄機(jī)組運(yùn)行穩(wěn)定性及控制其誘導(dǎo)噪聲輻射水平，本文對(duì)水泵水輪機(jī)內(nèi)部壓力脈動(dòng)特性及其引起的流動(dòng)誘導(dǎo)噪聲進(jìn)行深入研究，以期對(duì)抽水蓄能機(jī)組過(guò)渡過(guò)程運(yùn)行穩(wěn)定性提供一些技術(shù)指導(dǎo)。

國(guó)內(nèi)外學(xué)者針對(duì)水力機(jī)械內(nèi)部非定常壓力脈動(dòng)及其流動(dòng)誘導(dǎo)噪聲開(kāi)展了相關(guān)工作，Kuethe[10]在其專利中介紹了控制噴氣式發(fā)動(dòng)機(jī)、壓縮機(jī)、渦輪機(jī)等的噪音和不穩(wěn)定性；Kato等[11]將流體流動(dòng)和結(jié)構(gòu)的單向耦合模擬運(yùn)用到預(yù)測(cè)五級(jí)離心泵外表面輻射噪聲中；Jiang等[12]使用平行顯式動(dòng)態(tài)有限元模擬了泵結(jié)構(gòu)的振動(dòng)，闡明了共振噪聲產(chǎn)生和傳播的機(jī)制；Abbot等[13]通過(guò)分析一臺(tái)35 MW水泵水輪機(jī)氣蝕噪聲檢測(cè)結(jié)果，說(shuō)明了測(cè)量的噪聲水平應(yīng)與單元的空化侵蝕速率特性成比例，從而通過(guò)噪聲量確定空蝕損失；Yang等[14-15]結(jié)合計(jì)算流體力學(xué)與計(jì)算聲學(xué)對(duì)離心泵內(nèi)外場(chǎng)的噪聲進(jìn)行研究，得到截面區(qū)域聲壓峰值和壓力脈動(dòng)峰值位置對(duì)應(yīng)；Opperwall等[16]結(jié)合有限元與邊界元模型，研究了流體噪聲源對(duì)液壓泵和電機(jī)產(chǎn)生的空氣噪聲影響；王宏光等[17]模擬了軸流泵流場(chǎng)和聲場(chǎng)分布,并分析了有泵殼振動(dòng)影響和無(wú)泵殼振動(dòng)影響下泵殼體邊界聲場(chǎng)分布；鄭源等[18]采用間接邊界元法對(duì)由葉片旋轉(zhuǎn)偶極子源所引起的外場(chǎng)噪聲進(jìn)行數(shù)值計(jì)算；司喬瑞[19]分析了離心泵低噪聲在水力設(shè)計(jì)中可優(yōu)化項(xiàng)等。上述針對(duì)水力機(jī)械內(nèi)部壓力脈動(dòng)特性以及流動(dòng)誘導(dǎo)噪聲的研究主要集中于離心泵、軸流泵和混流泵，然而鮮有文章針對(duì)水泵水輪機(jī)壓力脈動(dòng)特性對(duì)流動(dòng)誘導(dǎo)噪聲的影響進(jìn)行研究，尤其是在過(guò)渡過(guò)程。

由于試驗(yàn)研究中對(duì)水泵水輪機(jī)整體內(nèi)部流場(chǎng)不能做到完全可視化，且在機(jī)組長(zhǎng)期運(yùn)行后，磨損使得導(dǎo)葉關(guān)閉狀態(tài)相鄰導(dǎo)葉體不能完全貼合接觸等原因，數(shù)值模擬方法成為研究水泵水輪機(jī)甩負(fù)荷過(guò)程的主要手段。此外，隨著高性能計(jì)算機(jī)的快速發(fā)展，計(jì)算流體力學(xué)技術(shù)被廣泛應(yīng)用于各行各業(yè)，數(shù)值計(jì)算模擬研究方法具有省時(shí)高效、節(jié)約成本、全可視化等優(yōu)勢(shì)，使得其成為當(dāng)下研究水力機(jī)械的主要手段，已有大量文獻(xiàn)表明CFD數(shù)值模擬研究方法的可靠性[19-21]。

綜上所述，本文將以水泵水輪甩負(fù)荷過(guò)程作為研究對(duì)象，采用CFD數(shù)值模擬方法分析其流場(chǎng)壓力脈動(dòng)特性，并以流場(chǎng)數(shù)值計(jì)算結(jié)果葉片表面壓力脈動(dòng)信號(hào)作為聲場(chǎng)流動(dòng)誘導(dǎo)噪聲的計(jì)算聲源，進(jìn)一步研究聲場(chǎng)噪聲輻射及指向性分布。

1 計(jì)算模型及方法

1.1 幾何模型和網(wǎng)格劃分

如圖1a所示為本文研究對(duì)象水泵水輪機(jī)三維模型圖，其包含進(jìn)水管道、帶11個(gè)固定導(dǎo)葉的前導(dǎo)葉域、帶22個(gè)導(dǎo)葉的活動(dòng)導(dǎo)葉域、帶7個(gè)后傾式三維葉片的轉(zhuǎn)輪域（轉(zhuǎn)輪進(jìn)口直徑=0.4 m）和出水管道，比轉(zhuǎn)速n=37.2。六面體結(jié)構(gòu)化網(wǎng)格被用于劃分整體流道，所有葉片區(qū)域均采用O-Block結(jié)構(gòu)，在ICEM中生成5套非結(jié)構(gòu)化網(wǎng)格，通過(guò)網(wǎng)格無(wú)關(guān)性檢驗(yàn)后發(fā)現(xiàn)當(dāng)總網(wǎng)格數(shù)大于1 200萬(wàn)時(shí)，所求得水頭波動(dòng)小于0.45%，為了捕捉到相關(guān)量的變化細(xì)節(jié)，最終劃分總網(wǎng)格數(shù)為1 400萬(wàn)，其中前導(dǎo)域的網(wǎng)格數(shù)500萬(wàn)，活動(dòng)導(dǎo)葉域395萬(wàn)，轉(zhuǎn)輪域373萬(wàn)[22]，圖1b所示為部分區(qū)域網(wǎng)格圖。

圖1 水泵水輪機(jī)模型與網(wǎng)格

1.2 湍流模型及其邊界條件

DES模型是對(duì)雷諾時(shí)均N-S方程（reynolds averaged navier stokes equations，RANS）模型的修改，其能夠?qū)τ诖鬁u模擬（large eddy simulation，LES）計(jì)算足夠精細(xì)的區(qū)域切換到亞格子尺度公式，然而對(duì)于靠近邊界區(qū)域與湍流長(zhǎng)度尺度小于最大網(wǎng)格尺寸的區(qū)域，均和RANS模型處理方式相同。因此，DES模型求解計(jì)算湍流流動(dòng)比LES模型需要的計(jì)算量少，但比RANS模型或者其他湍流模型更能準(zhǔn)確的捕捉流場(chǎng)細(xì)節(jié)[23]。DES湍流模型具體控制方程見(jiàn)文獻(xiàn)[24]，其已被成功運(yùn)用于一系列工程案例，結(jié)果表明該模型對(duì)流域存在顯著分離的模擬計(jì)算相對(duì)其他模型更為準(zhǔn)確[24-26]。因此本文采用計(jì)算中選用分離渦模擬（detached eddy simulation，DES）湍流模型[22]，借由商業(yè)軟件 ANSYS CFX 16.1對(duì)水泵水輪機(jī)甩負(fù)荷過(guò)程進(jìn)行數(shù)值模擬計(jì)算。

進(jìn)水管道進(jìn)口設(shè)置總壓（total pressure），出水管道出口設(shè)置為Opening（子選項(xiàng)選擇entrainment），流道壁面均設(shè)置為無(wú)滑移壁面，不同過(guò)流部件間設(shè)置交界面（GGI)，在動(dòng)靜過(guò)流部件之間，穩(wěn)態(tài)計(jì)算設(shè)置凍結(jié)轉(zhuǎn)子交界面（FRI），瞬態(tài)計(jì)算設(shè)置瞬態(tài)動(dòng)靜交界面（TR/SI）。時(shí)間步長(zhǎng)采用二階隱式離散，且步長(zhǎng)對(duì)應(yīng)轉(zhuǎn)輪旋轉(zhuǎn)1度，每步循環(huán)計(jì)算的最大次數(shù)為5，主要變量收斂階級(jí)為5階。同時(shí)，為保證準(zhǔn)確求解動(dòng)態(tài)特性，數(shù)值模擬計(jì)算中庫(kù)朗數(shù)要低于3。

模型試驗(yàn)在意大利帕多瓦大學(xué)工業(yè)工程系開(kāi)式旋轉(zhuǎn)機(jī)械試驗(yàn)臺(tái)（OTF）上完成，圖2a所示為試驗(yàn)臺(tái)結(jié)構(gòu)[27]，該試驗(yàn)臺(tái)按照國(guó)際標(biāo)準(zhǔn)IEC60041與IEC60193，針對(duì)水力機(jī)械性能及瞬態(tài)特性測(cè)試而設(shè)計(jì)模型驗(yàn)證結(jié)果如圖2b所示，基于ISO標(biāo)準(zhǔn)的定轉(zhuǎn)速工況系列計(jì)算結(jié)果與試驗(yàn)結(jié)果對(duì)比用于驗(yàn)證模型，相關(guān)量測(cè)量遵循IEC標(biāo)準(zhǔn)，儀器校準(zhǔn)均在現(xiàn)場(chǎng)完成[22]?？梢钥闯隹紤]流體弱可壓縮性得，到的計(jì)算結(jié)果與試驗(yàn)值更符合，水頭計(jì)算值與試驗(yàn)值吻合較好，設(shè)計(jì)工況附近誤差約為0.2%，部分負(fù)荷工況下誤差低于3.3%（圖2b），說(shuō)明本文所涉及的計(jì)算模型及方法可較為準(zhǔn)確的預(yù)測(cè)水泵水輪機(jī)的外特性，為下文研究水泵水輪機(jī)內(nèi)部壓力脈動(dòng)特性提供了保證。

圖2 模型試驗(yàn)

1.3 監(jiān)測(cè)點(diǎn)位置

導(dǎo)葉運(yùn)動(dòng)遵循如圖3所示的關(guān)閉規(guī)律，其中、是二次函數(shù)段是線性函數(shù)段，橫坐標(biāo)T代表相對(duì)時(shí)間，縱坐標(biāo)C表示相對(duì)導(dǎo)葉開(kāi)度。水泵水輪機(jī)甩負(fù)荷過(guò)程導(dǎo)葉運(yùn)動(dòng)引起網(wǎng)格扭曲變形，網(wǎng)格正交性降低導(dǎo)致數(shù)值模擬真實(shí)性降低，導(dǎo)葉大幅度運(yùn)動(dòng)產(chǎn)生負(fù)網(wǎng)格引起計(jì)算報(bào)錯(cuò)停止，若采用重畫網(wǎng)格等形式解決上述問(wèn)題，則導(dǎo)葉運(yùn)動(dòng)過(guò)程會(huì)出現(xiàn)間歇性，與電站實(shí)際情況不符。因此本文基于網(wǎng)格壁面滑行技術(shù)[22]，在整個(gè)計(jì)算過(guò)程中，為保證數(shù)值模擬的真實(shí)性，任一時(shí)刻網(wǎng)格質(zhì)量為初始網(wǎng)格質(zhì)量的60%以上。本文在進(jìn)行流動(dòng)誘導(dǎo)噪聲分析時(shí)，是通過(guò)非定常計(jì)算得到的壓力脈動(dòng)信號(hào)作為聲源。為更好的獲取水泵水輪機(jī)內(nèi)各處壓力脈動(dòng)信息，本文在模型內(nèi)設(shè)置了相應(yīng)的監(jiān)測(cè)點(diǎn)，部分監(jiān)測(cè)點(diǎn)位置如圖4所示，用于采集壓力信號(hào)以分析流場(chǎng)壓力脈動(dòng)特性，以及作為聲場(chǎng)流動(dòng)誘導(dǎo)噪聲研究的聲源。

圖3 導(dǎo)葉關(guān)閉規(guī)律

圖4 部分監(jiān)測(cè)點(diǎn)位置圖

1.4 聲場(chǎng)計(jì)算模型

聲學(xué)邊界元法基于波動(dòng)方程可以很好的對(duì)復(fù)雜腔體內(nèi)的聲場(chǎng)進(jìn)行數(shù)值計(jì)算，對(duì)低頻噪聲求解時(shí)具有明顯的優(yōu)勢(shì)，水泵水輪機(jī)在實(shí)際運(yùn)行過(guò)程中的聲學(xué)計(jì)算空間并非完全封閉，因此本文采用間接聲學(xué)邊界元法進(jìn)行噪聲數(shù)值計(jì)算，其是指將計(jì)算流體力學(xué)和計(jì)算聲學(xué)求解相結(jié)合，且對(duì)離散格式、湍流模型及邊界條件的要求均較低[18]，從本質(zhì)上講間接法是Lighthill聲類比法，控制方程如式（1）所示。

圖5 聲場(chǎng)殼體網(wǎng)格及監(jiān)測(cè)面布置

2 計(jì)算結(jié)果分析

2.1 流場(chǎng)計(jì)算結(jié)果分析

為了量化分析數(shù)值模擬計(jì)算得到壓力信號(hào)，定義壓力系數(shù)C如式（2）來(lái)表征壓力脈動(dòng)特性。

圖6所示為活動(dòng)導(dǎo)葉域進(jìn)口面監(jiān)測(cè)點(diǎn)GVI1、GVI8及GVI15在導(dǎo)葉關(guān)閉過(guò)程中壓力脈動(dòng)變化情況，可以看出各監(jiān)測(cè)點(diǎn)壓力系數(shù)具有一致的變化趨勢(shì)和近似相等的脈動(dòng)幅值范圍。導(dǎo)葉持續(xù)關(guān)閉使喉部直徑逐漸減小引起法向過(guò)流面積減小，導(dǎo)致部分水流被阻擋在導(dǎo)葉前而無(wú)法通過(guò)，前導(dǎo)葉域和活動(dòng)導(dǎo)葉域之間積聚大量的流體使得該無(wú)葉區(qū)壓力值呈現(xiàn)增長(zhǎng)趨勢(shì)（圖6a），且壓力脈動(dòng)幅值隨著時(shí)間減小。圖6b是圖6a在相對(duì)時(shí)間區(qū)間T=0.98到1之間的放大窗口，各監(jiān)測(cè)點(diǎn)壓力脈動(dòng)幅值范圍偏差說(shuō)明在接近導(dǎo)葉關(guān)閉點(diǎn)時(shí)，導(dǎo)葉進(jìn)口前無(wú)葉區(qū)內(nèi)流動(dòng)并非對(duì)稱。

圖6 活動(dòng)導(dǎo)葉進(jìn)口壓力脈動(dòng)特性

與導(dǎo)葉進(jìn)口處壓力變化趨勢(shì)相反，活動(dòng)導(dǎo)葉出口處壓力隨著時(shí)間減?。▓D7a），壓力脈動(dòng)幅值范圍在T=0.8之前呈現(xiàn)縮小趨勢(shì)，在最后20%導(dǎo)葉關(guān)閉階段（T=0.8到T=1）其幅值范圍持續(xù)增長(zhǎng)。從圖6和圖7可以看出導(dǎo)葉出口處壓力系數(shù)脈動(dòng)幅值范圍是對(duì)應(yīng)導(dǎo)葉進(jìn)口處的2倍以上，主要原因是轉(zhuǎn)輪旋轉(zhuǎn)使得該無(wú)葉區(qū)水流具有周向速度，另一方面是因?yàn)榛顒?dòng)導(dǎo)葉和轉(zhuǎn)輪之間產(chǎn)生較強(qiáng)的動(dòng)干涉影響流態(tài)。此外，導(dǎo)葉出口處壓力信號(hào)具有明顯的周期性，圖7b是圖7a在區(qū)間T=0.98到T=1之間的放大窗口，與圖6b不同之處在于各監(jiān)測(cè)點(diǎn)壓力脈動(dòng)幅值在相同范圍內(nèi)，垂直連續(xù)線段示意監(jiān)測(cè)點(diǎn)GVO1壓力周期特性。

圖7 活動(dòng)導(dǎo)葉出口壓力脈動(dòng)特性

圖8所示為無(wú)葉區(qū)內(nèi)壓力脈動(dòng)頻譜圖，可以看出2個(gè)無(wú)葉區(qū)內(nèi)壓力功率譜密度（power spectral density, PSD）主頻位置均在葉頻=0.051與=1處，且各監(jiān)測(cè)點(diǎn)在葉頻處所對(duì)應(yīng)頻譜幅值明顯高于其他頻率對(duì)應(yīng)頻譜值，從而說(shuō)明葉頻所對(duì)應(yīng)聲場(chǎng)的能量在流動(dòng)誘導(dǎo)噪聲頻譜中占主導(dǎo)作用，因此后文在對(duì)水泵水輪機(jī)進(jìn)行流動(dòng)誘導(dǎo)噪聲計(jì)算時(shí)將主要針對(duì)葉頻及其倍頻進(jìn)行求解。導(dǎo)葉出口處頻譜值是進(jìn)口處的10倍之上，由于活動(dòng)導(dǎo)葉進(jìn)口位置流態(tài)主要受動(dòng)靜干涉影響，而活動(dòng)導(dǎo)葉出口位置流態(tài)主要受動(dòng)動(dòng)干涉影響，說(shuō)明動(dòng)動(dòng)干涉對(duì)流場(chǎng)的影響要強(qiáng)于動(dòng)靜干涉。

轉(zhuǎn)輪出口流態(tài)相對(duì)復(fù)雜，監(jiān)測(cè)點(diǎn)壓力系數(shù)變化趨勢(shì)可分為3個(gè)階段：導(dǎo)葉關(guān)閉前50%（T=0到T=0.64）階段模型內(nèi)部流場(chǎng)相對(duì)穩(wěn)定，流量減少引起尾水管內(nèi)環(huán)繞內(nèi)壁面B逐漸形成空腔環(huán)域（圖9），轉(zhuǎn)輪出口壓力系數(shù)呈增長(zhǎng)趨勢(shì)（圖10a）；當(dāng)T=0.67時(shí)內(nèi)壁面B附近出現(xiàn)反向回流且其旋向與轉(zhuǎn)輪轉(zhuǎn)向相反（圖9），尾水管內(nèi)流道被分成3個(gè)區(qū)域（I-出流區(qū)，II-空腔區(qū)，III-回流區(qū)），同時(shí)反向回流區(qū)向尾水管外壁面A擴(kuò)散，至T=0.8尾水管內(nèi)空腔區(qū)域消失，因該階段尾水管內(nèi)存在兩個(gè)反向旋壁渦帶，監(jiān)測(cè)點(diǎn)DT1、DT2及DT3壓力脈動(dòng)加劇且幅值較大，表現(xiàn)出極不穩(wěn)定的脈動(dòng)特性（圖10a）；由于導(dǎo)葉關(guān)閉最后階段入流量小，回流區(qū)域延伸占據(jù)三分之二尾水管流道。為更清楚觀察尾水管進(jìn)口監(jiān)測(cè)點(diǎn)壓力脈動(dòng)情況，圖10b是圖10a區(qū)間T=0.98至T=1的放大圖。

圖8 無(wú)葉區(qū)壓力脈動(dòng)頻域特性

圖9 尾水管壓力分布及流線圖

圖10 轉(zhuǎn)輪出口壓力脈動(dòng)特性

2.2 聲場(chǎng)計(jì)算結(jié)果分析

水力機(jī)械流動(dòng)誘導(dǎo)噪聲指向性分布在各個(gè)方向上的趨勢(shì)基本相同，且其趨勢(shì)不隨流量的變化而改變[18]，因此選取面分析外場(chǎng)噪聲，研究一階到三階葉頻處噪聲指向性分布和輻射水平。從圖11a可以看出葉片表面偶極子聲源對(duì)應(yīng)的聲場(chǎng)云圖具有明顯的類似“∞”的指向形態(tài)，且各階葉頻處，聲壓的分布在面上呈現(xiàn)出明顯的對(duì)稱性，上述指向性分布具有對(duì)稱性證明葉片噪聲輻射具有明顯的偶極子特性。

注：BPF是葉頻，Hz。

圖11a所示為0.5BEP時(shí)一階葉頻聲壓云圖，可以看出聲壓擴(kuò)散形狀與模型結(jié)構(gòu)相關(guān)，在近似軸向流動(dòng)方向和近似軸向流動(dòng)的法向方向聲壓分布是對(duì)稱衰減的，且模型所在位置聲壓幅值最大，聲壓幅值隨著與模型距離的增加而減小。此外，最大聲壓值位于尾水管內(nèi)，圖11a中尾水管內(nèi)劍齒狀區(qū)域?qū)?yīng)圖9（T=0.64）出流區(qū)。

綜合分析圖11b、11c、11d不同流量下模型外聲場(chǎng)在一階、二階及三階葉頻處指向性分布，可以看出聲壓極小值位于0°和180°附近，極大值位于90°附近，且極大值的位置在尾水管附近，主要原因是尾水管內(nèi)水流一方面受到轉(zhuǎn)輪旋轉(zhuǎn)影響，具有一定的周向速度生成帶旋流，另一方面因?yàn)榱髁繙p小，尾水管內(nèi)水流不足以補(bǔ)充空腔位置，加劇了渦流的形成。圖8b所示一階葉頻輻射水平，其聲壓值明顯高于二階葉頻和三階葉頻，分析原因可能是殼體結(jié)構(gòu)的某階固有頻率與一階葉頻相近，導(dǎo)致流體與殼體之間產(chǎn)生共振，從而使得噪聲輻射水平較大。該現(xiàn)象從另一角度說(shuō)明流動(dòng)誘導(dǎo)噪聲的主，頻由壓力脈動(dòng)主頻和殼體固有頻率綜合決定，在設(shè)計(jì)模型時(shí)需重點(diǎn)考慮該問(wèn)題，盡可能的避免葉頻與殼體固有頻率接近。此外，流量從BEP減小至0.5BEP的過(guò)程中，一階葉頻聲壓級(jí)逐漸減?。欢S著導(dǎo)葉繼續(xù)關(guān)閉（0.5BEP至0.02BEP階段），聲壓級(jí)反而增長(zhǎng)。

噪聲在二階葉頻處輻射水平如圖11c所示，二階葉頻處不同流量下外聲場(chǎng)指向性分布與尾水管內(nèi)流場(chǎng)變化相互對(duì)應(yīng)，經(jīng)歷如下3個(gè)階段：流量從BEP減小至0.6BEP的過(guò)程中，二階葉頻聲壓級(jí)隨流量的減小而減小，聲壓幅值在流量大于0.5BEP變化較小，BEP處聲壓級(jí)是0.6BEP處的1.5倍；在流量為0.6BEP時(shí)，聲場(chǎng)分布在XY面上接近于圓形，說(shuō)明此時(shí)外場(chǎng)噪聲不同方位上的輻射能力差異不大，對(duì)應(yīng)于尾水管流場(chǎng)從0.6BEP至0.4BEP空腔區(qū)域形成回旋流（圖9），二階葉頻處聲壓級(jí)增大；在尾水管出流區(qū)和回流區(qū)空間縮小階段時(shí)，聲壓級(jí)減小，當(dāng)流量進(jìn)一步減小不足以填充該空環(huán)區(qū)時(shí)，聲壓級(jí)出現(xiàn)增長(zhǎng)趨勢(shì)（0.3BEP至0.02BEP）。圖11d中三階葉頻處流量從BEP減小至0.7BEP的過(guò)程聲壓級(jí)變化不大，導(dǎo)葉繼續(xù)關(guān)閉至0.5BEP階段聲壓級(jí)有明顯的降低，在后50%導(dǎo)葉關(guān)閉過(guò)程中，三階葉頻處聲壓級(jí)出現(xiàn)較為紊亂的發(fā)展趨勢(shì)。

3 結(jié) 論

本文基于DES湍流模型和動(dòng)網(wǎng)格理論對(duì)某電站可逆式水泵水輪機(jī)模型發(fā)電工況下導(dǎo)葉關(guān)閉過(guò)程進(jìn)行連續(xù)性模擬，通過(guò)分析流場(chǎng)與聲場(chǎng)計(jì)算結(jié)果得到以下結(jié)論：導(dǎo)葉進(jìn)出口處2個(gè)無(wú)葉區(qū)內(nèi)壓力脈動(dòng)主頻位置均在葉頻為斯特勞哈爾數(shù)等于0.051 與1處，且導(dǎo)葉出口處頻譜值是進(jìn)口處的10倍之上，說(shuō)明動(dòng)動(dòng)干涉對(duì)流場(chǎng)的影響要強(qiáng)于動(dòng)靜干涉。轉(zhuǎn)輪出口監(jiān)測(cè)點(diǎn)壓力脈動(dòng)特性與尾水管內(nèi)流場(chǎng)變化的3個(gè)階段相對(duì)應(yīng)（第一階段：相對(duì)時(shí)間T=0到T=0.64；第二階段：T=0.64到T=0.8；第三階段T=0.8到T=1），當(dāng)尾水管內(nèi)出現(xiàn)2個(gè)反向旋壁渦帶時(shí)（T=0.67），壓力脈動(dòng)最強(qiáng)烈且其幅值最大。在導(dǎo)葉持續(xù)關(guān)閉的過(guò)程中，外場(chǎng)噪聲的主頻由壓力脈動(dòng)主頻與殼體固有頻率綜合決定，且各階（1、2、3階）葉頻處聲壓指向性分布和輻射水平分別反映出流場(chǎng)壓力脈動(dòng)特性。葉片表面偶極子聲源對(duì)應(yīng)的聲場(chǎng)云圖具有明顯的類似“∞”的指向形態(tài)，且各階葉頻處聲壓的分布呈現(xiàn)出明顯的對(duì)稱性，說(shuō)明葉片噪聲輻射具有明顯的偶極子特性。在1階、2階葉頻處，流動(dòng)誘導(dǎo)噪聲聲壓分布整體表現(xiàn)為：導(dǎo)葉關(guān)閉前50%階段，流量對(duì)于外場(chǎng)噪聲輻射能力的影響表現(xiàn)為大流量工況下較強(qiáng)（設(shè)計(jì)工況流量BEP到0.6BEP），小流量工況下較弱（0.6BEP到0.4BEP），導(dǎo)葉關(guān)閉后50%階段正好相反，從而說(shuō)明改善水泵水輪機(jī)內(nèi)部壓力脈動(dòng)情況是降低流動(dòng)誘導(dǎo)噪聲輻射水平的重要手段。

[1] 李君，王磊，廖偉麗. 可逆式水泵水輪機(jī)“S”形區(qū)域內(nèi)部流場(chǎng)特性分析[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2014，30(15)：106－113.

Li Jun, Wang Lei, Liao Weili. Internal flow characteristics analysis on S-shaped region of reversible pump-turbine[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2014, 30(15): 106－113. (in Chinese with English abstract)

[2] Deane J P, Gallachóir B P ó, Mckeogh E J. Techno- economic review of existing and new pumped hydro energy storage plant[J]. Renewable & Sustainable Energy Reviews, 2010, 14(4): 1293－1302.

[3] Sun H, Xiao R, Liu W, et al. Analysis of S characteristics and pressure pulsations in a pump-turbine with misaligned guide vanes[J]. Journal of Fluids Engineering, 2013, 135(5): 511011.

[4] Liu J, Liu S, Wu Y, et al. Numerical investigation of the hump characteristic of a pump–turbine based on an improved cavitation model[J]. Computers & Fluids, 2012, 68: 105－111.

[5] Mao X, Monte A D, Benini E, et al. Numerical study on the internal flow field of a reversible turbine during continuous guide vane closing[J]. Energies, 2017, 10(7): 988.

[6] Li D, Gong R, Wang H, et al. Analysis of vorticity dynamics for hump characteristics of a pump turbine model[J]. Journal of Mechanical Science & Technology, 2016, 30(8): 3641－3650.

[7] Liu D M, Zheng J S, Wen G Z, et al. Numerical simulation on the “S” characteristics and pressure fluctuation of reduced pump-turbine at start-up condition[J]. 2012, 15(6): 062034.

[8] Hasmatuchi V, Roth S, Botero F, et al. High-speed flow visualization in a pump-turbine under off-design operating conditions[C]// 25th IAHR Symposium on Hydraulic Machinery and Systems, 2010: 012059.

[9] Xiao Y X, Sun D G, Wang Z W, et al. Numerical analysis of unsteady flow behaviour and pressure pulsation in pump turbine with misaligned guide vanes[C]// 26th IAHR Symposium on Hydraulic Machinery and Systems. 2012, v 15, n PART3.

[10] Kuethe A. Control of noise and instabilities in jet engines, compressors, turbines, heat exchange and the like：US3776363 [P/OL]. (1973-04-12) [2018-06-30]. http://www.freepatentsonline.com/3776363.html.

[11] Kato C, Yoshimura S, Yamade Y, et al. Prediction of the noise from a multi-stage centrifugal pump[C]// ASME 2005 Fluids Engineering Division Summer Meeting, 2005: 1273－1280.

[12] Jiang Y Y, Yoshimura S, Imai R, et al. Quantitative evaluation of flow-induced structural vibration and noise in turbomachinery by full-scale weakly coupled simulation[J]. Journal of Fluids & Structures, 2007, 23(4): 531－544.

[13] Abbot P A, Walsh J, Halas R. Cavitation noise investigation of a pump-turbine[C]// Waterpower. ASCE, 2015: 2031－2040.

[14] Yang J, Yuan S, Yuan J, et al. Numerical and experimental study on flow-induced noise at blade-passing frequency in centrifugal pumps[J]. Chinese Journal of Mechanical Engineering, 2014, 27(3): 606－614.

[15] 袁壽其，薛菲，袁建平，等. 離心泵壓力脈動(dòng)對(duì)流動(dòng)噪聲影響的試驗(yàn)研究[J]. 排灌機(jī)械，2009，27(5)：287－290.

Yuan Shouqi, Xue Fei, Yuan Jianping, et al. Experimental study on impact of pressure fluctuation on flow-noise in centrifugal pump[J]. Drainage and Irrigation Machinery, 2009, 27(5): 287－290. (in Chinese with English abstract)

[16] Opperwall T, Vacca A. A combined FEM/BEM model and experimental investigation into the effects of fluid-borne noise sources on the air-borne noise generated by hydraulic pumps and motors[J]. Proceedings of the Institution of Mechanical Engineers Part C Journal of Mechanical Engineering Science, 2014, 228(3): 457－471.

[17] 王宏光，徐小龍，楊愛(ài)玲，等. 軸流泵流動(dòng)噪聲數(shù)值模擬[J]. 排灌機(jī)械工程學(xué)報(bào)，2011，29(3)：199－203.

Wang Hongguang, Xu Xiaolong, Yang Ailing, et al. Numerical simulation of flow noise in axial-flow pump[J]. Drainage and Irrigation Machinery, 2011, 29(3): 199－203. (in Chinese with English abstract)

[18] 鄭源，陳宇杰，毛秀麗，等. 混流泵壓力脈動(dòng)特性及其對(duì)流動(dòng)誘導(dǎo)噪聲的影響[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2015，31(23)：67－73.

Zheng Yuan, Chen Yujie, Mao Xiuli, et al. Pressure pulsation characteristics and its impact on flow-induced noise in mixed-flow pump[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2015, 31(23): 67－73. (in Chinese with English abstract)

[19] 司喬瑞. 離心泵低噪聲水力設(shè)計(jì)及動(dòng)靜干涉機(jī)理研究研究[D]. 鎮(zhèn)江：江蘇大學(xué)，2014.

Si Qiaorui. Investigation on Hydraulic Design of Centrifugal Pumps with Low Noise and Mechanism of Rotor-Stator Interaction[D]. Zhenjiang: Jiangsu University, 2014. (in Chinese with English abstract)

[20] Zhao X, Xiao Y, Wang Z, et al. Numerical analysis of non-axisymmetric flow characteristic for a pump-turbine impeller at pump off-design condition[J]. Renewable Energy, 2018, 115: 1075－1085.

[21] Li D, Wang H, Qin Y, et al. Numerical simulation of hysteresis characteristic in the hump region of a pump-turbine model[J]. Renewable Energy, 2018, 115: 433－447.

[22] Mao X, Giorgio P, Zheng Y. Francis-type reversible turbine field investigation during fast closure of wicket gates[J]. Journal of Fluids Engineering, 2018, 140(6): 061103.

[23] Spalart P R. Detached-eddy simulation[J]. Annual Review of Fluid Mechanics, 2009, 41(1): 181－202.

[24] Minakov A V, Engineer D V P, Litvinover I V, et al. Vortex ropes in draft tube of a laboratory Kaplan hydroturbine at low load: an experimental and LES scrutiny of RANS and DES computational models[J]. Journal of Hydraulic Research, 2017(1): 1－18.

[25] Xu H, Wray T, Agarwal R K. Application of a new DES model based on wray-agarwal turbulence model for simulation of wall-bounded flows with separation[C]// Aiaa Fluid Dynamics Conference, 2017.

[26] Hui S, Xiao R, Wang F, et al. Analysis of the pump-turbine S characteristics using the detached eddy simulation method[J]. Chinese Journal of Mechanical Engineering, 2015, 28(1): 115－122.

[27] 陽(yáng)君，袁壽其，Pavesi Giorgio，等. 水泵水輪機(jī)泵工況下近設(shè)計(jì)點(diǎn)駝峰現(xiàn)象的流動(dòng)機(jī)理研究[J]. 機(jī)械工程學(xué)報(bào)，2016，52(24)：170－178.

Yang Jun, Yuan Shouqi, Pavesi Giorgio, et al. Study of hump instability phenomena in pump turbine at large partial flow conditions on pump mode[J].Journal of Mechanical Engineering, 2016, 52(24): 170－178. (in Chinese with English abstract)

Simulation of flow induced noise in process of pump-turbine load rejection

Mao Xiuli1, Sun Aoran2, Giorgio Pavesi3, Zheng Yuan4, Ge Xinfeng4

(1.712100; 2.230088,; 3.35131,; 4.210000)

The grid market is redistributed with significantly increase of the exploitation of unpredictable renewable energy, such as wind and solar energy sources, however, the ways of electricity generation by both wind and solar energy depend on environment which is extremely unstable. For the sake of balancing electricity generated by renewable energy, pumped storage power stations are experiencing a thriving process. As the core of pumped storage power station, the stable operation of the pump turbine is extremely important, especially for pump-turbine working at transient conditions. In order to study pressure fluctuating characteristics and its' influence on flow-induced noise, a continuous unsteady simulation was carried out in pump-turbine guide vane closing process under generating mode. In this article, wall sliding mesh was used to realize guide vane continuous motion, which ensured that the mesh quality at any moment was larger than 60% compared with the initial mesh quality, meanwhile, DES turbulent model was adopted in all calculations due to its good performance in many industrial cases. The whole pump-turbine model was meshed with structured mesh by commercial software ICEM, and five different mesh sizes were used in mesh sensitivity validation, with the size of 14 million selected finally. On the other hand, a test was performed by the team of Giorgio Pavesi to prove this model in open test facility in Padova University, the entire model validation was carried out according to ISO standards, and relative parameters were measured based on IEC standards. Commercial software ANYSYS CFX 16.2 was used to realize all simulating calculations with 8 computer cores, one month was taken to finish this calculation. The flow field calculating results were analyzed in frequency and time-frequency domains, including mass flow, pressure, and torque et al., in the meantime, the pressure on the surfaces of blades was regarded as flow-induced noise source to study sound field. The solution obtained from flow field illustrates that pressure fluctuating amplitudes at guide vane outlet is more than twice compared to the relative value at guide vane inlet location, the main reason is flow in the vaneless space that is close to runner is affected by rotor-rotor interaction. In addition, pressure pulsations at runner outlet arrive at peaks when two vortexes appear in draft tube with two different rotating directions. As for frequency domain characteristics, both strauhal number=0.051 and=1 are captured, whereas the spectrum of those pressure fluctuations that are close to guide vane outlet is 10 times of the relative value at guide vane inlet, which explains that rotor-rotor interaction has a stronger influence on flow field than rotor-stator interaction. Some rules are found by analyzing flow-induced noise in sound field, the analysis illustrates that flow-induced noise radiation level is related to both pressure fluctuating and shell natural frequency captured in exterior acoustic field, the shape of sound distribution is like “∞” and sound level distributions in different directions and faces are symmetrical, this explains that the blade noise radiation has obvious dipole characteristics. Furthermore, at the first and second-order blade passage frequencies, the effect of flow rate on the radiation performance of noise is stronger under larger flow conditions during guide vane closure, which becomes weaker under smaller flow conditions in the first half of the guide vane closure, as for the second half phase of guide vane closure, the results are exactly opposite to the previous phenomena. Moreover, flow-induced noise radiation is consistent with fluid characteristics during pump-turbine load rejection. Consequently, to improve pressure fluctuating characteristics can reduce flow-induced noise.

pressure; impellers; noises; pump-turbine; load rejection process

10.11975/j.issn.1002-6819.2018.20.007

TK734

A

1002-6819(2018)-20-0052-07

2018-04-17

2018-07-30

國(guó)家公派留學(xué)項(xiàng)目（No. 201506710011）

毛秀麗，講師，從事水力機(jī)械及其系統(tǒng)研究。Email：maoxl@nwafu.edu.cn

毛秀麗，孫奧冉，Giorgio Pavesi，鄭 源，葛新峰. 水泵水輪機(jī)甩負(fù)荷過(guò)程流動(dòng)誘導(dǎo)噪聲數(shù)值模擬[J]. 農(nóng)業(yè)工程學(xué)報(bào)，2018，34(20)：52－58. doi：10.11975/j.issn.1002-6819.2018.20.007 http://www.tcsae.org

Mao Xiuli, Sun Aoran, Giorgio Pavesi, Zheng Yuan, Ge Xinfeng. Simulation of flow induced noise in process of pump-turbine load rejection[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2018, 34(20): 52－58. (in Chinese with English abstract) doi：10.11975/j.issn.1002-6819.2018.20.007 http://www.tcsae.org