姓名：劉舒
學號：309005736
呈交日期：02/05/2011
Name:    Shu LIU
Student Number: 309005736
Date Submitted: 02/05/2011
 
Pipe Flow and Head Loss due to Friction管道流量和水龍頭的摩擦損失

合法的合作可以被定義為任何有建設性的教育和提高智力的做法，其目的是通過學生之間的互動，以促進最佳的學習成果。這些做法的典型例子包括：
Legitimate co-operation can be defined as any constructive educational and intellectual practice that aims to facilitate optimal learning outcomes through interaction between students. Typical examples of these practices may include:
研究和寫作的相關項目/工程/教科文件
• the researching and writing of joint projects/written works/tutorial papers;
普通的主題和概念的討論
• discussion of general themes and concepts;
評估準則的解釋
• interpretation of assessment criteria;
非正式學習/小組討論
• informal study/discussion groups;
通過互相之間的協作能力鞏固和發展的學術寫作技巧
• strengthening and development of academic writing skills through peer assistance;
合法的合作是基于生產工作是由獨立的學生（或一組學生而進行一個聯合項目）的責任的原則，同時認可學生之間的相互合作的價值。
Legitimate co-operation is based on the principle that producing the work remains the independent responsibility of the student (or group of students where a joint project is undertaken), while recognising the educational value of interaction between students.

土木工程學院的學術誠信表可從土木工程網站下載
The School of Civil Engineering Statement on Academic Honesty is available from the Civil Engineering Website:
 
Academic Honesty Declaration
By submitting this report electronically I am confirming that my assignment meets the academic honesty requirements
·         This assessment is my own work
·         I understand the requirements for academic honesty
·         I am aware of plagiarism and its penalties
·         I have not re-used previously submitted material in this assignment
·         I have not engaged someone else to complete this assignment
·         I have retained a duplicate hard copy of this assignment
Quality assurance Declaration
·         I have proof read my report#p#分頁標題#e#
·         I have spell checked my report
·         I have read the notes on writing a report. Report Tips  and  Report Rules(delete two of the headings below)

Introduction
There are two different head losses when water flow though a pipe. One is static head loss, which is caused by elevation. The other one is dynamic head loss, since water will lose potential energy due to the friction between pipe and water. It is a potential energy lost, instead of dynamic energy lost, since the flow rate is identical alone the pipe. There are basically two different types of flow, laminar flow and turbulent flow, which can be determined by Reynold’s number.
As McKEON AND SHARMA (2010) pointed out, ‘The accurate descriptions of statistical scaling and instantaneous structural coherence of turbulent fluctuations, and their relationship to the mean flow, are amongst the important unsolved problems in physics. Even at the simplest level, we still lack a complete explanation of the development of the mean velocity profile in canonical flows as the Reynold’s number increases.’
The first part of the experiment was done to found out the relationships between static head loss and elevation. The second part of the experiment was also done to find out the roughness, represented as friction factor, of a smooth pipe in both laminar flow and turbulent flow. The experiment results were compared with theoretical values. Exact values were hard to be predicted when Reynold’s number is large during the dynamic head loss test.

Methods
Static head loss test and dynamic head loss test were done in this experiment.

Static Head Loss Test:
The flow started at a high flow rate. The free jet, P, at each tapping was measured (Image 1). The pipe was titled and the free jet P’ at each tap was measured again. The height of each tap above the horizontal, Z, was also measured (Image 2).

 Dynamic head loss Test
A manometer tube was connected to three taps on a flow pipe (Image 3). This is read directly to give the head at each tapping. The manometer reads the head. (Head is the height equivalent to pressure per unit mass) To read the head for the laminar flow use the micrometer insert to work out the head loss.

Image 3: Diagram of flow in pipe with three taps connected to a manometer.

The experiment will be conducted 8 times in total, twice at each different flow rate:
Measure and record the distances between the taps on the pipe, L1-2 and L2-3 (refer to Image 3 above). Also record the internal diameter of the pipe. Take the water temperature and use the appropriate graph to determine, the kinematic viscosity (m2/s). Bleed the manometer by opening all 4 valves at once and then close all valves.
The following steps were repeated for each trial: The collection box was drained. Close the drain. Set the balance to zero and start the stopwatch at the same time. After 20kg water was collected, the stopwatch was stopped. Above steps were repeated twice more to get 3 readings. Open each manometer valve separately to observe the head at positions 1, 2 and 3 and record as h1, h2 and h3.#p#分頁標題#e#
When going down from high flow rate to medium flow rate, leave tap 1 open and reduce the flow so that the head drops about 30%. Apply a similar reduction when going down from medium to low flow rate. For the laminar flow decrease the flow rate and attach the nozzle to the end of the pipe. Use the measuring cylinder rather than the collection box and time how long it takes to fill the cylinder. To determine fFDS, the theoretical friction factor read from the Moody chart, use the curve for smooth pipes.

Static Head Loss
For static head loss, it can be see clearly from figure 1 that the free jet when tilted plus the height of tap is almost exactly the same as free jet when the pipe is horizontal. This proves that the static head loss at the same distance from entrance of pipe is caused by elevation of the pipe.
On the other hand, the head loss between tap 1 and tap 2 and the head lose between tap 2 and tap 3 is different in both horizontal and tilted pipes. This head loss should be caused by friction and also influenced by the free end of the pipe. This head loss should be considered as dynamic head loss. There are two different types of streamlines for the free end and the free jet. The distance between free end and free jet and flow rate determine the friction which, in turn, affects the pressure distribution. It will strike a balance and the head loss alone the pipe is not linear. However, the exact relationship need further experiment to be found out.
There could be large error, approximately 0.01m, when measuring the height of the tap when titled. Since the two distances between the three taps are same, the differences between the heights of the three taps when titled should also be same. In experiment, however, the height differences are 0.17m and 0.05m which suggest that the pipe is not perfect straight. This may causes errors when measuring the actual height free jet in static head loss test.

Dynamic Head Loss
The pipe is assumed to be smooth pipe. Figure 2 use log axis for flow rate which is linear to Reynolds Number for the same pipe, so the shapes of the curves are similar to the shape of the smooth pipe curve in Moody Chart. Figure 2 clearly shows that the experimental results of Laminar flow, Low Flow Rate and High Flow Rate are pretty close (in error bar) to the theoretical friction values. This proves that when the flow rate is not high (Reynolds Number is less than approximate 30000), the dynamic head loss is caused by the friction between water and pipe. The High Flow Rate tests (Reynold’s Number is greater than approximate 30000), however, gave much smaller friction values than theoretical ones. It is a big sudden drop which should be caused by other factors, such as significant turbulence in flow. It can be seem from figure 3 that the assumption of smooth pipe may not be valid, because the curve of relative roughness of 0.0002 could be chosen to give more accurate predicts.
Absolute roughness depends on the type and age of the material of pipe, while relative roughness also depends on the diameter of the pipe. In assumption, the pipe is smooth pipe which has zero absolute and relative roughness. In theory, a new drawn copper pipe should have absolute roughness, e, of 1.5 x 10-6m. In experiment, the absolute roughness of pipe is 3.2 x 10-6m, which indicates that the pipe is an old pipe.#p#分頁標題#e#
For a particular pipe, Re is linear to the flow rate. Laminar flow occurs when a fluid flows in parallel layers, with no disruption between the layers (Batchelor, 2000). Laminar flow usually has low flow rate which gives low Re. Turbulent flow usually occurs when flow rate is high and has many unsteady vortices which appear on many scales and interact with each other. For pipe flow, a Reynolds number above about 4000 will most likely correspond to turbulent flow, while a Reynold's number below 2100 indicates laminar flow. The region in between (2100 < Re < 4000) is called the transition region (Fox, Robert W. 2009). In laminar flow, f=64/Re is used to predict the friction factor, while in turbulent flow, the curves in the Moody chart are used to predict the friction factor.
In this experiment, there is no test fall in the transition zone where more and more parts of the flow begin to change from laminar flow to turbulent as the flow rate increases. In transition zone, the linear interpolation could be used to predict the Re, although there could be large error.
When measuring the head at three tap, there is sometimes few air bubbles in the plastic pipe. This may causes error when reading the height. In the two high flow rate tests, 40kg water should be collected, instead of 20kg, to give more accurate flow rate.

Conclusion
Tow experiments were done to find what factors cause the static head loss and dynamic head loss. Static head loss is caused by elevation of the pipe. Dynamic head loss is caused by friction and should be considered as potential energy loss of flow. The assumption that the pipe is smooth is not valid. The copper pipe has relative roughness of 0.0002 and absolute roughness of 3.2 x 10-6m. The pipe is old while a new drawn copper pipe should have absolute roughness, e, of 1.5 x 10-6m. When Reynold’s number is less than 2100, the flow is laminar and experimental results agree with the theoretical result. When Reynold’s number is greater than 4000, the flow becomes turbulent, but the curve of relative roughness 0.0002 in Moody Chart still provides accurate predicts of friction factor until Reynold’s number exceeds 30000. When the flow rate is very high, Reynold’s number greater than 30000, the experiment results gives much smaller friction factor than that in theory. The reason could be the significant turbulence in the flow which makes the assumption of parabolic curve of speed distribution not valid.

Reference
Batchelor, G.K. 1973. An Introduction to Fluid Dynamics. Cambridge University Press
Fox, Robert W. 2009. Introduction to fluid mechanics. 7th Edition. John Wiley. Hoboken, N.J. http://ukthesis.org/xxhzsglzy/
McKEON, B. J. A. S. SHARMA and J. Fluid Mech(Editor). 2010. A critical-layer framework for turbulent pipe flow. vol. 658, pp. 336–382. Cambridge University Press