Flow phenomena of a siphonic roof outlet arc Buitenhuis Abstract Introduction

Transcription

1 Flow phenomena a siphonicc ro outlet Marc Buitenhuis Hydraulic research engineer Akatherm BV, Panningen, Netherlands Abstract So far the investigations on siphonic rof drainage systems have been focused on the functioning the system as a whole. In this report the focus will be on the ro outlet and the phenomena observed o during tests in a test rig according the EN 1253 and ASME A112 standardss for ro outlets. This comprises the test a single ro outlet in a siphonic configuration. primary goal the testss was design a well-functioning all plastic improved ro outlet, which is currently c in production. During the testss some remarkable phenomena typicall siphonic ro outlets were w observed, that will be presented in this report. It appeared that air ingress follows a path decreasing size air pockets (plugs bubbles), followed by a decreasee in number bubbles. noise level produced by the ro outlet is correlated with w this phenomenon. Further it was observed thatt a disc alone is not sufficient keep the air out the system and that additional measures, like fins on p the air baffle, are necessary prevent the development large vortices that are capable dragging air in the system. Also it was observed that waves reflect from the edges the disc and fins the t air baffle leading an interferencee pattern waves around the ro outlet. It was observed that the air is still being sucked s in from a certain discharge ratee after the water level hass reached the height the disc the air baffle. At first large plugss air are being dragged in, but with increasing water level the plugs become bubbles decreasing in size and increasing in number. noise level appearss be correlated this phenomenon. Finally a pumping behavior was observed at the outlet when having becomee siphonic that is probablyy correlated instabilities the flow at the exit the pipe system. 1. Introduction Siphonic ro drainage systems are gaining more and more ground in the drainage rain water from large flat ros because their cost effective, space saving design and high performance. Significant work has been done by several institutes describe the principles the system. design the siphonic ro outlet on p this system has not been investigated this intensively thus far. refore tests have been performed with our in- ro outlet design leading the development an all house test rig gain insight in the principles the plastic siphonic ro outlet for ro drainage systems exceeding all requirements the EN and ANSI standards and the build up a vast knowledge the functioning a siphonic ro outlet. tests havee been performed in a single ro outlet siphonic configuration. On p a special component has been developed use the ro outlet in ann emergency ro outlet configuration as well without a significant decreasee in performance compared c the standard ro outlet. In this paper the t phenomena observed at the siphonic ro outlet will be describedd one by one. First all the siphonic ro drainage system s itself and its components will be described in paragraph 2. n in paragraph 3 the basic working principle will be described. Following in paragraph 4 the flow phenomena occurring at the ro outlett will be described.

2 ro outlets form the entrance from the ro the pipe system. T tail pipes guide the water down the collecr pipe, which is not onlyy collecting the water from all ro outlets but also guides it horizontally o one side a building,, where the fall pipe transportss it downward wards the sewer. ro outlet consists a ro plate [A], a so called air baffle [B] and a a connecr attach it the pipe system [C] andd possibly a leaf guard [D] (which in our design has been integratedd in the air baffle design [B]). Illustration 3 presents the ro outlet and its components. Illustration 1. Test rig 2. Siphonic ro components drainage system and A siphonic ro drainage systems consists one or more ro outlets [1] followed by an equal amount vertical tail pipes [2] which leads the flow in a horizontal collecr pipe [3] that ends in a vertical fall pipe [4] (see illustration 2). As stated before the tests have been performed on a single ro outlet configuration. emergency ro outlet has also been tested as a single ro outlet configuration, thus not t in conjunction with a regular ro outlet system. its Illustration 3. Siphonic ro outlet ro plate is a speciallyy designed plate with a hole with smoothly rounded edges, which iss the opening the pipe system. air baffle is mounted over the opening in the ro platee prevent air from being drained in the pipe system (as will be further explained in a following paragraph). connecr is mounted on the botm part the ro plate connect the ro plate the pipe system. function the leafguard is keepp out rough sized s debris that can possibly block the system. 3. Workingg principle system the ro r drainage Illustration 2. Siphonic ro drainage system siphonic ro drainage system iss based on the principle full bore flow. This means that the system is designed o obtain a flow through the system that fills the whole diameter the pipe, thereby creating a vacuum upstream dragging more water in from the ro, thereby reaching r a high discharge rate through a relatively small diameter pipe system. To obtain a full bore flow the water must t be forced t fill the whole diameter, which can be done by decelerating it, e.g. in a horizontal collecr pipe, which leads a so called hydraulic jump, that can force the flow fill the whole diameter the pipe (see illustration 4).

3 4. Flow phenomena att the ro outlet ro outlet must prevent air from entering the siphonic ro drainage d system initiate full bore flow while having as a less resistance the system as possible obtain a maximum m discharge flow and minimum water height on o the ro Air ingress Illustration 4. Working principle siphonic ro drainage system hydraulic jump is like the piling up cars inn a traffic jam after a slow turn in the highway. water curtain at the end the tail pipe and the flow through the elbows before the fall pipe is characterized byy a ventilated mixture water and air as long as the flow has not becomee siphonic. gravity driven flow in the fall pipe accelerates the flow, and after becoming siphonic drags the water down functioning as the engine the system. So the water falling down the tail pipe, accelerating on its way down, is being decelerated in the collecr pipe make it flow full bore when enough water is discharged through the pipe system and is being accelerated in the fall pipe. When the hydraulic jump is able fill the pipe diameter the water will try force out the air in front against the upward buoyancy forces the air while filling the tail pipe upstream at the same time. This process is called priming. system is already functioning as a siphonic s system when about 60% the system is filled with water. air is still present then, but the system is no longer ventilated from p botm and thus the water falling down creates an underpressure, sucking s in more water from the ro. siphonic ro drainage system must be designed such that a once a year rain srm can be transported from the ro before the ro construction c is endangered. This is done by a proper choice pipe diameters in the system. This also means that the system will work conservatively (carrying air) for most regular rainfall events and siphonically on some cases heavy rainfall. In case more extreme rain srms an emergency ro drainage system must come in playy drain the excess rainwater up a maximum sum falling during a once in a 100 year rain srm. exact numbers for rainfall intensity for and capacity f a regular and emergency system and the allowable water levels on the ro can differ from countryy country. When there would w not be an air baffle, the water that is being dragged down thee pipe system would drag air along at the water w surface. water willl accelerate wards the ro outlet. velocity component the surrounding water will increase with the decreasee the radius around the Q v = 2π R h wate ro outlet ( er ). Following Bernoulli s laww the increasing velocity wards the pipe system will mean the waterheight will be decreasing wards the pipe p v + g h = const (since ρ 2 and thus h will decrease withh increasing v under the condition atmospheric pressure at the surface). water flowing wards the pipe system will drag air along at its surface, dragging it in the pipe system. Illustration 5. Air being b dragged inn with the flow o water In theory this would w be a symmetric situation occurring equally fromm all directions. In practice small disturbances (both( in thee construction the ro outlet, water flow and in environmental air flow) will lead a swirling flow wards the pipe system (vortex creation) similar as the vortex seen when emptying a bathtub.

4 4.2 Air baffle design To prevent air from being dragged in a disc could be mounted on p the pipe system. This will sp the air from being dragged in with the water flow. Unfortunately this will not be sufficient, since the water will still swirl around the disc and create a vortex that drags in air underneath the disc. A vortex accelerates the flow with a maximumm in its center, creating an under pressure. By Bernoulli s Law again this means the water surface is lowered, which can result in a local dip that creates the opportunity for air find its way underneath the ro outlet. Fins are necessary break the vortex formation up in smaller vortices that are not strong enough drag in air underneath the disc Phenomena typical for siphonic ro r outlets Using a disc shaped air baffle design with fins leads some specific phenomena. p As soon as thee water level around the air baffle reaches the disc the air baffle (note that the water level is decreasing wards the airr baffle and thus the overall water level is higher) the water will close the path for air enter the system. Due under pressures increasing with the discharge rate however, from a certain discharge rate the air will be sucked in. At first this is by largee plugs air, but when the water level rises they become bubbless decreasing in size with the water level. Att a further increase the water level the number bubbles will decrease. air being sucked in will collect underneath u the disc the air baffle from where it will be sucked in o the pipe system. collision o water on the edge the t disc and fins the air baffle will lead reflection waves. This is seen as an interference pattern waves around the ro outlet, which becomes more significant at higher discharge rates. above phenomena aree illustrated in the illustration below (illustration 7). (a) no disc or air baffle, large vortex Illustration 7. Floww phenomena at t ro outlet (b) disc, vortex at edge disc (ink injected) (c) disc with fins, no vortex (ink injected) Illustration 6. Ro outlet with no disc (a), with discc (b), with disc with fins (c) As long as thee water level is below the lower edgee the disc the air will be dragged in with the water as in a conventional system, the only influence the air baffle being the preventionn a large vortex v by the fin shaped ribs onn which the baffle stands. When the water level reaches the edgee the disc the air can not freely enter thee system anymore but has be sucked in by the under pressure generated by the water flow. Att a low discharge rate the flow velocity is o low drag in air, but when the water velocity increases largee plugs air manage pass underneath the disc (illustration 8a). Initially the water is only occasionally capable closing the gap for air enter underneath the disc. When the water level increases the plugs get shorter until bubbles remain (illustration 8b). air has overcome an increasing resistance by the water find

5 its way in when the water takes in more and more space above the air baffle. When the discharge rate increases further the size the bubbles decreases while the amount bubbles increases (illustration 8c). (a) large plugs air just after reaching closuree water surface at the disc the air baffle diminishes as the water level rises above the air baffle disc) and full bore b flow is a fact. When the floww has become fully siphonic a wave train is seen at the water surfacee around the air baffle every s (thus at a frequency Hz). It looks like the ro outlett is pumpingg out water at this frequency producing a wave w train with every pump stroke. In reality the water intake will be restricted every 0.75 s, probably because a pressure wave traveling upstream fromm the exit. At full siphonic flow the 75 ro outlet (with butt welded pipes) transports approximatelyy 16.5 l/s, which converts a mean m velocityy 4.4 m/s in the pipe. Since the pipe system is 4.22 m (tal fall height) +1.5 m (length horizontal collecr pipe) = 5.7 m long and the wave frequency is Hz the speed at which the wave travels up and down the system iss 5.7*2*1.333 = 15.2 m/s, which is approximately 3.5 times the velocity through the pipe. Indeed the exit flow is not a straight jet, it wobbles from left right, although the frequency reported above can not be confirmed. swinging motion appears occur at a frequency 8 Hzz (=6*1.333 Hz)! It is not clear how explain the facr 6 between the wobbling the exit flow and the wave frequency at the ro outlet Noise levels (b) large air bubbles due higher water levell breaking up the plugs Along with the break up in smaller plugs and bubbles the noise production rises as the noise is a result the opening and closing c thee water surface on the edge the disc, which occurs more frequently when bubbles are dragged in at a high frequency. When the frequency thee bubbles diminishes so does the noise production (because the fewer opening and closing incidents thee water surface on the edge the disc, seee illustration 9 below). (c) small bubbles at increased water level, siphonic flow, inducing a high noise level but well before Illustration 8. Air ingress underneath air baffle (transparent) When the water level rises further the frequency the bubbles diminishes until the air is no longer capable getting underneath the disc (the air is no longer capable breaking the water surface find its wayy in while the under pressure at the water surface

6 noise curve plugss siphonic 80 noise (db) bubbles No air ingress 8 10 discharge (l/s) Illustration 9. Typical noise curve siphonic ro outlet 5. Conclusions In So far the investigations on siphonic ro drainagee systems have been focused on the functioning f the system as a whole. In this report the focus will be on the rof outlet and the phenomena observed during tests in a test rig according the EN 1253 and ASME A112 standardss for ro outlets. This comprises the test a single ro outlet in a siphonic configuration. primary goal the tests was design a well-functioning typical siphonic ro outlets were observed, that will be presented in all plastic improved ro outlet, which is currently in production. During the tests some remarkable phenomena this report. It appeared that air ingress follows a path decreasing size air pockets (plugs bubbles), followed by a decrease in number bubbles. noise level produced by the ro outlet is correlated with w this phenomenon. Further it was observed that a disc alone is not sufficient keep the airr out the system and that additional measures, like fins on p the air baffle, are necessary prevent the development large vortices that t are capable dragging air in the system. Also it was observed that waves reflect from f the edges the disc and fins f the air baffle leading an interference pattern waves around the ro outlet. It was observed that the air is still being sucked in from a certain discharge rate after the water level has reached the height the disc the air baffle. At first large plugs air are being dragged in, but with increasing water level the plugs become bubbles decreasing in size and increasing in number. noise level appears be correlated this phenomenon. Finally a pumping behavior was observed at the outlet when having become siphonic that is probably correlated instabilities the flow at the exit the pipe system.