|
|
外文翻译
液压冲击原理
我们在分析液压冲击现象和合理的流体方程之前，首先先来描绘一般的关于压力传递的机械理论。通过参与这个关于阀门定位在一个较长点几乎没有摩擦的管道传输液体于两个蓄能源之间的结果之后是必要的。这个阀门连接的顺流管道截面和逆流管道截面考虑是一样的。压力冲击流将通过阀门操作传递在两个管道之间，并且假设阀门的关闭速度不应用于坚固圆管理论。
如果阀门是关闭的，而液体的流向是逆方向的，缓慢前进，结果导致液体被压缩和管道的横截面膨胀。阀门的压力增加导致高压液体逆向流动，延长了液体流过圆管通向阀这段管道的时间。这种高压液体的流动类似声音的传播，是依靠液体和管道材料作为介质的。
同样，阀的顺流面流动的延迟，将导致减小压力在阀门处。这个结果否定了高压液体的流动是沿着顺流管道的，阻止液体流动，假设流体压力在顺流管道是不能减小液体压力的或者蒸汽压力或者溶解气体释放的压力，各种愿意的考虑是不同的。
这样，关闭着的阀门导致高压液体的流动是沿着管道的，尽管那些流动有着各种不同的征兆。相对于稳定的压力流经阀门开启的管道。这种影响是关于液体流动的延迟在两种管道截面之间，管道自身受到影响由于液体逆向产生高压，管壁膨胀。同时，顺流管道缩短，由于流经液体的压力降低，这种管道横截面的巨大变形是由于管道材料的，并且能够被证明。例如，使用薄壁型橡胶管材。高压液体沿着液流前进。实践证明，由于液体的张力流向沿着管壁，它的速度接近于声速。在这种管道材料中，然而，这是一种次要作用，当认识到它的存在，能够解释一部分压力的传递时间随着阀门关闭特点，它几乎没有影响到压力标准应用在压力冲击现象。
在阀门关闭之后，这时是受压时间将主要依靠系统的边界条件，为了描绘阀门关闭的结果在同一个系统上，它将很容易说明在大量的图表上面，管道在每个时间段的情形。
由于阀门的关闭是瞬时的，液体接近每一段管道的阀门会带来停止，并且高压液体流动情况可能已经流过每一段管道。在适应的流速c和一段时间t，这时液体已经流过了一段距离1=ct，在每一段管道内，这时管道的横截面是变形也有一段距离1。
高压液体到达蓄能站通过管道的时间为t=1/c，在这段距离中出现了一个不稳定的位置，是在管道与蓄能站连接处。由于是不可能出现层流在蓄能站连接处，而保持压力不同及其它的值在阀门关闭之前，流过每一个蓄能站的时间为1/c，在逆向管道这边是高压液体的流动朝向阀门的关闭。减小管壁的压力到其原值，并且恢复管壁的横截面积。这时液体的流动需要产生差值。从管道流向蓄能站，在管道的前段的液体流动有比较高的压力比蓄能站。现在，由于系统假设没有摩擦，这种巨大的逆向流动会有精确的对比和最初的流动速度。
在顺流蓄能站，存在相反的情况，导致液体压力上升流向和确定的顺流流向从蓄能站到阀门。
由于这里考虑的是简单的管道，恢复高压液体在管道和阀门之间的时间为21/c。整个逆流管道也是同样，在返回最初的压力和流向在管道外也被确定时间为21/c，由于液体已经到达阀门，意味着没有液体提前在提供的逆向一个低的压力区域形成在阀门外，破坏了流向和给上升的压力减小流动流向逆方向的阀门。再一次，带来流动的停止沿着管道且减小压力在管道中。它已经被假设在阀门处压力下降，减小蒸发压力。由于系统已经假设没有摩擦，所有的液面会有相同，绝对的，巨大的压力增加。在稳定的运动压力下，会通过阀门的关闭产生。如果压力增长是h，这时所有的液面是h，因此，液体逆流经过阀门的时间为21/c，存在一个值-h,同时，减少所有沿着管道的点从h降到最初的压力时间逆向流动到蓄能站的时间为31/c。
类似的，恢复液体最初的顺流到阀门的时间为21/c，并且流向从顺流管道流向阀门关闭，这会在阀门处带来流动停止，导致压力上升。在整个顺流管道的每一段时间内压力h上升到最初的压力在流动停止时。
因此，在31/c时是一种不稳定的情形类似于在t=1/c的情形，出现在蓄能站和管道的连接处存在着不同。即是逆流管道压力下降到最初压力和顺流管道上升到最初压力，然而，这种液体流动恢复机构所用时间是相同的t=1/c。结果是逆流流向蓄能站，它有效地恢复环境沿着管道到它的最初值。当液体到达关闭的阀门时，沿着每一段管壁都是相同的时间t=0，然而，由于阀门一直是关闭的，这种情形不能保持循环流动周期。
管道系统采用循环流动周期，瞬时选择一种专门的机械情形，管道的顺流和逆流对于阀是一样的。实际 ，这是不同的。因而，所描绘的周期将一直被使用，除了压力变化在两管道之间不再表示相同相位关系，每一个压力周期的变化将是41/c，那里1和c代表着每一段管道适应的时期，这是重要的标记，一旦阀门是关闭的，这两个管道将做出相应的流动到任何一段距离。
通过上述冲击周期的描绘，可以划分压力-时间关系，在某一点沿管道上，这些变化的出现是类似的。通过时间在任何一点h，液体到达某一点，系统假设流动速度为一个常数c，这主要集中在压力冲击依靠的方法是限制压力的升高和减小阀的启闭速度。然而，存在着很重要的一点，没有减小开启压力，将发生直到阀的关闭时间先于另一个管道。减小压力达到出现阀门慢速关闭的结果先于忽略液体逆流到阀门关闭。由于没有影响，返回到阀门时间21/c前，从阀门开始运动没有压力减小能够到达如果阀门没有打开超过了时间。一般来说，阀门的关闭小于管道涉及的速度并且它将比21/c短。
在没有摩擦的情况下，周期的继续是不确定的。然而，实际中，摩擦力是压力损失在很短的时间内，系统的摩擦损失越高，忽略摩擦力的影响导致结果越严重。事实上，阀门的顶点低相对于蓄能站顶点。然而，由于缓慢的流动，摩擦点的损失减少。沿着管壁并且这个点向着蓄能站的方向增长。由于液体的每一层，在阀门和蓄能站中会带来停止，通过流动最初的液面，所以大多在第二个液面位置相应的摩擦点恢复流向。阀门导致影响整个时间21/c。由于流动是相反的在管道中时间为21/c和41/c。这个位置影响主要在阀门，由于重新建立一个新的摩擦损失，在确切的事例中，例如，长距离油管，在阀门关闭之前，它将上升一部分压力。
随着假设条件对摩擦周期的描绘，提及到使压力下降的条件，如果这些情况发生，这时流向圆管已经分离出类似的周期描绘，可能中断通过形成蒸气压力减小的位置有蒸气生成。因此，系统描述可能发生在阀门的顺流时间0或者逆流时间21/c形成一个腔。由于一段时间液体沿管壁流动在一个压力梯度下，在这个腔和系统边界之间。这种方法是通常由于产生额外压力在最后的腔中。这种现象一般涉及到像圆管的分离和通常的制作更多的错综复杂的由于释放溶解的气体在附近的腔中。
冲击压力也许被定义为在一些封闭的管道中应用，通过两个基本的方程，分别是运动平衡方程和连续应用在一个短的流体圆管。它依靠可变的流体平均压力和速度在任何一段管道的横截面，且不依靠可变的时间和距离。通常考虑实际的稳流方向。摩擦力将被假设与速度平方成比例，并且稳流摩擦关系将被假设应用在非稳定事例中。
Hydraulic transient theory
Before we embarking on the analysis of pressure transient phenomena and the derivation of the appropriate wave equations,it will be usefull to describe the general mechanism of pressure propagation by reference to the events fllowing the instantaneous closure of a value postioned at the med-length point of a frictionless pipeline carrying fluid between two reservoirs.The two pipeline sections upstream and downstream of the value are identical in all respects.Transient pressure waves will be propagated in both pipes by valve operation and it will be assumed that rate of value closure precludes the use of rigid column theory.
As the valve is closed,so the fluide approaching its upstream face is retarded with a consequent compression of the flude and an expansion od the pipe cross-section.The increase in pressure at the valve results in a pressure wave being propagated upstream which conveys the retardation of flow to the column of fluid approaching the valve along the upstream pipeline.This pressure wave travels through the fluid at the appropriate sonic velocity,which will be shown to depend on the properties of the fluid and the pipe material.
Similarly,on the downstream side of the valve the retardation of flow results in a reduction in pressure at the valve,with the result that a negative pressure waves is propagated along the downstream pipe which,in turn,retards the fluid flow.It will be assumed that this pressure drop in the downstream pipe is insufficient to reduce the fluid pressure to either its vapour pressure or its dissolved gas release pressure,which may be considerable different.
Thus,closure of the valve results in propagation of pressure waves along both pipes and,although these waves are of different sign relative to the steady pressure in the pipe prior to valve operation,the effect is to retard the flow in both pipe sections.The pipe itself is affected by the wave propagation as the upstream pipe swells as the pressure rise wave passes along it,while the downstream pipe contracts due to the passage of the pressure reducting wave.The magnitude of the deformation of the pipe cross-section depends on the pipe material and can be well demonstrated if,for example,thin-walled rubber tubing is employed.The passage of the pressure wave through the fluid is preceded,in practice,by a strain wave propagating along the pipe wall at a velocity close to the sonic velocity in the pipe material.However,this is a secondary effect and,while knowledge of its existence can explain some parts of a pressure-time trace following valve closure,it has little effect on the pressure levels generated in practical transient situations.
Following valve closure,the subsequent pressure-time history will depend on the conditions prevailing at the boundaries of the system.In order to describe the events following valve closure in the simple pipe system outlined above,it will be easier to refer to a series of diagrams illustrating conditions in the pipe at a number of time steps.
Assuming that valve closure was instantaneous,the fluid adjacent to the valve in each pipe would have been brought to rest and pressure waves conveying this information would have been propagated at each pipe at the appropriate sonic velocity c.At a later time t,the situation is as shown in fig.The wavefronts having moved a distance 1=ct,in each pipe,the deformation of the pipe cross-section will also have traveled a distancel as shown.
The pressure waves reach the reservoirs terminating the pipes at a time t=1/c.at this instant,an unbalanced situation arises at the pipe-reservior junction,as it is clearly impossible for the layer of fluid adjacent to the reservoir inlet to maintain a pressure different to that prevailing at that depth in the reservoir.Hence,a restoring pressure wave having a magnitude suffcient to bring the pipeline pressure back to its value prior to valve closure is transmitted from each reservoit at a time 1/c.For the upstream pipe,this means that a pressure wave is propagated towards the closed valve,reducing the pipe pressure to its original value and restoring the pipe cross-section.The propagation of this wave also preduces a fluid flow from the pipe into the reservoir as the pipe ahead of the moving wave is at a higher pressure than the reservoir.Now,as the system is assumed to be frictionless,the magnitude of this reversed flow will be the exact opposite of the original flow velocity,as shown in fig.
At the downstream reservoir,the converse occurs,resulting in the propagation of a pressure rise wave towards the valve and the establishment of a flow from the downstream reservoir towards the valve.
For the simple pipe considered here,the restoring pressure waves in both pipes reach the valve at a time 21/c.The whole of the upstream pipe has,thus,been returned to its original pressure and a flow has been established out of the pipe.At time 21/c,as the wave has reached the valve,there remains no fluid ahead of the wave to support the reversed flow.A low pressure region,therefore,forms at the valve,destroying the flow and giving rise to a pressure reducing wave which is transmitted upstream from the valve,once again bringing the flow to rest along the pipe and reducing the pressure within the pipe .It is assumed that the pressure drop at the valve is insufficient to reduce the pressure to the fluid vapour pressure.As the system has been assumed to be frictionless,all the waves will have the same absolute magnitude and will be equal to the pressure increment,above steady running pressure,generated by the closure of the valve.If this pressure increment is h,then all the waves propagating will be±h,Thus,the wave propagation upstream from the valve at time 21/c has a value-h,and reduces all points along the pipe to –h below the initial pressure by the time it reachs the upstream reservoir at time 31/c.
Similarly,the restoring wave from the downstream reservoir that reached the valve at time 21/c had established a reversed flow along the downstream pipe towards the closed valve .This is brought to rest at the valve,with a consequent rise in pressure which is transmitted.downstream as a +h wave arriving at the downstream reservoir at 31/c,at which time the whole of the downstream pipe is at pressure +h above the initial pressure whth the fuid at rest.
Thus,at time 31/c an unbalanced situation similar to the situation at t=1/c again arises at the reservoir –pipe junctions with the difference that it is the upstream pipe which is at a pressure below the reservoir pressure and the downstream pipe that is above reservoir pressure .However,the mechanism of restoring wave propagation is identical with that at t=1/c,resulting in a-h wave being transmitted from the upstream reservior,which effectively restores conditions along the pipe to their initial state,and a+h wave being propagated upstream from the downstream reservoir,which establishes a flow out of the downstream pipe.Thus,at time t=41/c when these waves reach the closed valve,the conditions along both pipes are identical to the conditions at t=0,i.e.the instant of valve closure.However ,as the valve is still shut,the established flow cannot be maintained and the cycle described above repeats.
The pipe system chosen to illustrate the cycle of transient propagation was a special case as,for convenience,the pipes upstream and downstream of the valve were identical.In practice,this would be unusual.However,the cycle described would still apply,except that the pressure variations in the two pipes would no longer show the same phase relationship.The period of each individual pressure cycle would be 41/c,where I and c took the appropriate values for each pipe.It is important to note that once the valve is closed the two pipes will respond separately to any further transient propagation.
The period of the pressure cycle described is 41/c.However,a term ofen met in transient analysis is pipe period,this is defined as the time taken for a restoring reflection to arrive at the source of the initial transient propagation and,thus,has a value 21/c.In the case described,the pipe period for both pipes was the same and was the time taken for the reflection of the transient wave propagated by valve from the reservoirs.
From the description of the transient cycle above,it is possible to draw the pressure-time records at points along the pipeline.These variations are arrived at simply by calculating the time at which any one of the±h waves reaches a point in the system assuming a constant propagation velocity c.The major interest in pressure transients lies in methods of limiting excessive pressure rises and one obcious method is to reduce valve speeds.However,reference to fig.illustrates an important point no reduction in generated pressure will occur until the valve closing time exceeds one pipe period.The reduction in peak pressure achieved by slowing the valve before a time 21/c from the start of valve closure and,as no beneficial pressure relief can be achieved if the valve is not open beyond this time.Generally,valve closures in less than a pipe period are referred to as rapid and those taking longer than 21/c are slow.
In the absence of friction , the cycle would continue indefinitely .However ,in practice, friction damps the pressure oscillations within a short period of time .In system where the frictional losses are high,the neglect of frictional effects can result in a serious underestimate of the pressure rise following valve closure.In these case,the head at the valve is considerably lower than the reservoir head.However,as the flow is retarded,so the frictional head loss is reduced along the pipe and the head at the valve increase towards the reservoir value.As each layer of fluid between the valve and the reservoir is brought to rest by the passage of the initial +h wave so a series of secondary positive waves each of a magnitude corresponding to the friction head recoverd is transmitted toward the valve,resulting in the full effect being felt at time 21/c.As the flow reverses in the pipe during time 21/c to 41/c,the opposite effect is recorded at the valve because of the re-establishment of a high friction loss,these variations being shown by lines AB and CD.In certain cases,such as long distance oilpipelines,this effect may contribute the larger part of the pressure rise following valve closure.
In addition to the assumptions made with regard to friction in the cycle description,mention was also made of the condition that the pressure drop waves at no time reduced the pressure in the system to the fluid vapour pressure.If this had occurred,then the fluid column would have separated and the simple cycle described would have been disrupted by the formation of a vapour cavity at the position where the pressure was reduced to vapour level.In the system described,this could happen on the valve’s downstream face at time 0 or on the upstream face at time 21/c.The formation of such a cavity is followed by a period of time when the fluid column moves under the influence of the pressure gradients between the cavity and the system boundaries.The period is normally terminated by the generation of excessive pressure on the final collapse of the cavity.This phenomena is generally referred to as column separation and is frequently made more complex by the release of dissolved gas in the vicinity of the cavity.
