- •Chapter I introduction
- •1. The subject of hydraulics
- •2. Historical background
- •3. Forces acting on a fluid. Pressure
- •4. Properties of liquids
- •Chapter II hydrostatics.
- •5. Hydrostatic pressure
- •6. The basic hydrostatic equation
- •7. Pressure head. Vacuum. Pressure measurement
- •8. Fluid pressure on a plane surface
- •Fig. 12. Pressure distribution on a rectangular wall
- •9. Fluid pressure on cylindrical and spherical surfaces. Buoyancy and floatation
- •Fig. 18. Automatic relief valve.
- •Relative rest of a liquid
- •10. Basic concepts
- •11. Liquid in a vessel moving with uniform acceleration in a straight line
- •12. Liquid in a uniformly rotating vessel
- •The basic equations of hydraulics
- •13. Fundamental concepts
- •14. Rate of discharge. Equation of continuity
- •15. Bernoulli's equation for a stream tube of an ideal liquid
- •16. Bernoulli's equation for real flow
- •17. Mead losses (general considerations)
- •18. Examples of application of bernoulli's equation to engineering problems
- •Chapter V flow through pipes. Hydrodynamic similarity
- •19. Flow through pipes
- •20. Hydrodynamic similarity
- •21. Cavitati0n
- •Chapter VI laminar flow
- •22.Laminar flow in circular pipes
- •23. Entrance conditions in laminar flow. The α coefficient
- •24. Laminar flow between parallel boundaries
- •Chapter VII turbulent flow
- •25. Turbulent flow in smooth pipes
- •26. Turbulent flow in rough pipes
- •27. Turbulent flow in noncircular pipes
- •Chapter VIII local features and minor losses
- •28. General considerations concerning local features in pipes
- •29. Abrupt expansion
- •30. Gradual expansion
- •31. Pipe contraction
- •32. Pipe bends
- •33. Local disturbances in laminar flow
- •34. Local features in aircraft hydraulic systems
- •Chapter IX flow through orifices, tubes and nozzles
- •35. Sharp-edged orifice in thin wall
- •36. Suppressed contraction. Submerged jet
- •37. Flow through tubes and nozzles
- •38. Discharge with varying head (emptying of vessels)
- •39. Injectors
- •Relative motion and unsteady pipe flow
- •40. Bernoulli's equation for relative motion
- •41. Unsteady flow through pipes
- •42. Water hammer in pipes
- •Chapter XI calculation of pipelines
- •43. Plain pipeline
- •44. Siphon
- •45. Compound pipes in series and in parallel
- •46. Calculation of branching and composite pipelines
- •47. Pipeline with pump
- •Chapter XII centrifugal pumps
- •48. General concepts
- •49. The basic equation for centrifugal pumps
- •50. Characteristics of ideal pump. Degree of reaction
- •51. Impeller with finite number of vanes
- •52. Hydraulic losses in pump. Plotting rated characteristic curve
- •53. Pump efficiency
- •54. Similarity formulas
- •55. Specific speed and its relation to impeller geometry
- •56. Relation between specific speed and efficiency
- •57. Cavitation conditions for centrifugal pumps (according to s.S. Rudnev)
- •58. Calculation of volute casing
- •59. Selection of pump type. Special features of centrifugal pumps used in aeronautical and rocket engineering
Chapter VII turbulent flow
25. Turbulent flow in smooth pipes
It was mentioned in Sec. 19 that turbulent flow is characterised by a mixing of the fluid and fluctuations of velocity and pressure. A sensitive recorder of velocity fluctuations would draw a graph like the one in Fig. 52. The magnitude of the velocity varies erratically about a certain mean value which is constant for a given condition.
The pathlines traced by fluid particles through a fixed point in space are curves of different shape, despite the straightness of the pipe. The streamlines also vary greatly at any given instant (Fig. 53). Thus, strictly speaking, turbulent flow is unsteady as the velocity, pressure intensity and pathlines of the fluid particles change with time.
Nevertheless, turbulent flow can be regarded as steady if the mean velocity and pressure and the total discharge do not change with time. This kind of flow is frequently encountered in engineering practice.
The velocity distribution (averaged over a period of time) ^across a section of turbulent flow differs markedly from velocity distribution in laminar flow.
A comparison of the velocity profiles in the same pipe and with the same discharge (same mean; velocity) in conditions of laminar and turbulent flow reveals a pronounced difference (Fig. 54). Velocity distribution in turbulent flow is somewhat more uniform and the increase in velocity at the boundary is steeper than in laminar flow, in which, it will be recalled, the velocity profile is a parabola.
Consequently, the a coefficient, which takes account of nonuni-form velocity distribution in Bernoulli's equation (see Sec. 16), is much smaller in turbulent flow. Unlike laminar flow, where a does not depend on Re (see Sec. 23), in turbulent flow a is a function of Re, decreasing with the latter increasing from 1.13 at Re = Rer to 1.025 at Re = 3 X 10е. As the graph in Pig. 55 shows, thecurve of a as a function of Re approaches unity asymptotically. In most cases of turbulent flow a can be assumed to be unity.
Energy losses in turbulent flowthrough pipes of uniform cross-section (i. e., head losses due to friction) also differ from the case of laminar flow. In the former friction losses are much greater for the same rate of discharge, viscosity and pipe size.
This increased loss is due to the formation of eddies, mixing and curving of pathlines. Whereas in laminar flow friction losses increase as the first power of the velocity (and the discharge), in transition to turbulent flow a jump is observed in the resistance and then a steeper increase of hf along a curve approaching a parabola of the second power (Fig. 56).
In view of the complexity of turbulent flow and the difficulty of analytical study, the theory of turbulent flow lacks precision. The so-called semiempirical, approximate turbulence theories of Frandol, von Karman and others are applied, but we shall not consider them here.
In most cases engineering calculations of turbulent flow in pipos are based on purely experimental data systematised on the basis of hydrodynamic similarity.
The fundamental equation for turbulent flow in circular pipes is the experimental formula (4.18), which follows directly from similarity considerations and has the form
or
where λt = friction factor for turbulent flow.
This basic formula is valid for both turbulent and laminar flow (see Sec 22), the difference lying in the value of λ.
As in turbulent flow friction losses vary approximately as the square of the velocity (and the square of the rate of discharge), the friction factor in Eq. (4.18) can, to a first approximation, be assumed constant.
However, it follows from the similarity law (Sec. 20) that λt, like λt should be a function of the fundamental criterion of similarity, the Reynolds number, which includes velocity, diameter and kinematic viscosity, i. e.,
There are a number of empirical and semiempirical formulas which express this function for turbulent flow through smooth pipes;
one of the most convenient and commonly used is a formula developed by the Russian scientist P. K. Konakov:
(7.1)
which holds good for values ranging from Re = Recr to several milions.
When 2,300 < Re < 105 the old Blasius equation can be used:
(7.2)
It will be observed that with Re increasing Kt decreases somewhat, though to a much smaller degree than in laminar flow (Fig. 57).
This difference in the change of X is due to the fact that the effect of viscosity on friction is much less in turbulent than in laminar flow. Whereas in the latter friction losses vary directly as the viscosity (see Sec. 22), in turbulent flow, as is evident from Eqs (4.18) and (7.2), they vary as the l/4th power of viscosity. This is because of mixing and momentum transport in turbulent flow.
Eqs (7.1) and (7.2) for determining the friction factor Xt in terms of Re are valid for so-called technically smooth pipes, the roughness of which is small enough so as practically not to affect resistance. To a sufficient degree of accuracy, as technically smooth can be regarded drawn nonferrous (including aluminium alloy) tubing and special seamless high-grade steel pipes. Thus-, the pipes used in aircraft fuel systems and hydraulic transmissions can be regarded as smooth and the foregoing equations used in calculating them. Since steel water pipes and cast iron pipes offer greater resistance, they cannot be treated as smooth, and Eqs (7.1) and (7.2) are not valid for them.
The resistance of rough pipes will be examined later on.
It is known from similarity theory, and confirmed by experiments carried out by a number of researchers (Nikuradse, G. G. Gurzhiyenko and others) that in turbulent flow a laminar sublayer usually forms at the pipe walls (Fig. 58). This is a very thin layer in which the fluid moves much slower and without mixing.
Within the sublayer the velocity increases steeply from zero at the wall to a certain value vt at the sublayer boundary. The sublayer is very thin; the Reynolds number computed for 6P the velocity vt and the kinematic viscosity v, was found to be a constant quantity:
(7.3)
where δ = sublayer thickness.
This quantity is a universal constant, like the critical Reynolds number for flow through pipes. When the velocity of motion and, consequently, the Reynolds number increase, the velocity vt increases as well and the sublayer thickness 6, decreases. At large values of Re the laminar sublayer practically disappears.