The Herschel Venturi Lab

AE: 341 Fluid Mechanics Laboratory.

Section 6

Author Wyatt Welch

Instructor Mohamed Amine Abassi

September 22, 2024

The Herschel Venturi Lab Wyatt Welch 2

Introduction

The Herschel Venturi Lab studies the change of pressure through a pipe’s throat, and the left-over

effects. The objectives of this lab will be to determine the coefficient of discharge as a function of

Reynolds Number, and to compare the experimental data with accepted empirical results. We will

also determine overall head loss as a function of maximum head differential and compare with

accepted empirical results. This experiment was performed by Wyatt Welch and fellow classmates,

on SDSU’s campus, room EIS 010. This lab experiment and report was a requirement for AE341.

This experiment will be conducted by controlling the water through a pipe. This pipe has a 130

mm section which undergoes diameter shrinking, a period of consistent throat size, then expanding

to the initial diameter. The test will be measuring the state of flow at the beginning, throat, and

ending of this section (labeled sections 1, T, and 2, respectively). At these points, there are holes

that allow water to flow upwards from pressure. The flow rate is controlled to specific points using

valves, and we measure the volume and time of water passing through the system at given

pressures. With this data we use formulas derived from the class to interpret learned concepts.

Theory

Utilizing Bernoulli’s Equation, and assuming no head loss through the pipe and that this can be

considered 1-D flow, we can derive changes in pressure and velocity across the system. On top of

that, we use the Continuity Equation to derive a change in velocity formula between section 1, and

section T with respect to a change in diameter. This derivation gives us our first equation, which

calculates the theoretical Velocity at T using the changes in height of tubes and the diameter ratio.

2𝑔𝑔∆ℎ1−𝑇𝑇 𝑚𝑚
𝑉𝑉𝑇𝑇 = � 𝑑𝑑 ( ) equation 1
1−(𝐷𝐷)4 𝑠𝑠
The Herschel Venturi Lab Wyatt Welch 3

From equation 1, we use Volume Flow Rate to solve for 𝑄𝑄𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 , by multiplying the equation

by the area of its cross section.

2𝑔𝑔∆ℎ1−𝑇𝑇 𝑚𝑚3
𝑄𝑄𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 = 𝑉𝑉𝑇𝑇 𝐴𝐴 𝑇𝑇 = 𝐴𝐴 𝑇𝑇 � 𝑑𝑑 ( ) equation 2
1−(𝐷𝐷)4 𝑠𝑠

𝑄𝑄𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 can be solved in a simpler method, using the measured results from the experiment

we can divide the volume by its recorded time for the pipe to produce that volume. And since we

took 2 recordings at each test, we average the two to get the average Volume Flow Rate for each

test. The coefficient of discharge is the ratio between the experimental and theoretical volume flow

rates and should be between 0 and 1. The closer to 1, the more accurate it is.

𝑉𝑉𝑉𝑉𝑉𝑉𝑢𝑢𝑢𝑢𝑢𝑢 𝑚𝑚3
𝑄𝑄𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 = ( ) equation 3
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑠𝑠

𝑄𝑄𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
𝐶𝐶 = equation 4
𝑄𝑄𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒

The Reynold’s number is the ratio of inertial to viscous fluids in a system. This was given for the

class. This involves density (ρ), characteristic velocity (C*), characteristic length (L), and dynamic

viscosity (µ) which was found using a textbook table.

𝜌𝜌𝐶𝐶 ∗ 𝐿𝐿
𝑅𝑅𝑅𝑅 = equation 5
𝜇𝜇

∆ℎ1−𝑇𝑇 and ∆ℎ1−2 are found using the instrument the experiment was preformed on. Between

each test, the flow rate was calibrated such that ∆ℎ1−𝑇𝑇 would decrease by 50 mm, starting from

400 mm. Once that had been set, ∆ℎ1−2 was observed and recorded.
The Herschel Venturi Lab Wyatt Welch 4

Results and discussion

From our calculations, accepted empirical data for Re and C are provided in figure 1. This was

used to plot a standard expected result. To pair with that, all data recorded from the experiment is

provided in figure 2.

Re C
2000 0.91
5000 0.94
10000 0.955
20000 0.965
50000 0.97
Figure 1: Empirical data for Re and C

run Volume Time 𝑸𝑸𝒂𝒂𝒂𝒂𝒂𝒂 ∆𝒉𝒉𝟏𝟏−𝑻𝑻 ∆𝒉𝒉𝟏𝟏−𝟐𝟐 Re C

# Litres sec 𝟑𝟑 mm mm
𝒎𝒎 /𝒔𝒔
1 25 35 42.04 58.77 0.595107 400 105 27767.03 0.983793
2 25 35 45.25 63.07 0.553713 350 80 25835.62 0.978564
3 25 35 48.93 67.82 0.513503 300 75 23959.49 0.980214
4 15 25 32.24 53.7 0.465405 250 60 21715.28 0.973195
5 15 25 36.51 60.61 0.41166 200 45 19207.59 0.962414
6 15 25 42.07 69.57 0.357949 150 30 16701.53 0.966306
7 15 25 51.09 84.75 0.294292 100 20 13731.36 0.973011
8 15 25 72.23 120.25 0.207785 50 10 9695.024 0.971556
Figure 2: Collected data, and calculated Re, C, and 𝑄𝑄𝑎𝑎𝑎𝑎𝑎𝑎

From the data gathered and shown here, we can follow the calculations that will be covered in

the sample calculation section of this report. First, we find 𝑄𝑄𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 by calculating the

average between the two measurements in a run. Next, we calculate velocity using that number

and dividing it by the area in section 1. Using that, as well as equation 5, we find Reynold’s

number. We calculate 𝑄𝑄𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 with the help of equation 2, and the combination of these two

figures gives us C via equation 4. Below are two plots showing the data gathered from the lab.
The Herschel Venturi Lab Wyatt Welch 5

Figure 3: Experimental and empirical data calculated from experimentation

Figure 4: Relation between the change in distance between section 1-T, and 1-2
The Herschel Venturi Lab Wyatt Welch 6

The experimental data in figure 3 may seem odd, and there are certainly some outliers that make

the data less reliable, however overall, there is a trend that can be compared to the empirical data.

It appears there is a slight logarithmic trend which correlates to the empirical. Minus two data

points, one of which being an especially egregious outlier, the trend is clear. Although, C is high

all throughout the test, which would suggest a strong accuracy in the results. Figure 4 has a much

stronger relationship, with a correlation coefficient of 0.9935. This trend shows that there is a

relationship between the change in pressure from sections 1-T, and 1-2.

Conclusion

The data presented in this lab is consistent with the expected results that were discussed in the

lecture, as well as what we would expect conceptually. Sources of error are most likely to arise

from troubles balancing the heights of the tubes, as they were constantly fluctuating. Besides that,

small errors would have certainly accumulated from recordings of time and litre counts, as well as

from adjustments to flow speeds. The figures presented are all within acceptable ranges of error

and do a good job of displaying what could be expected from seeing changes in tube shapes in the

real world. Thanks to the work of Herschel, and his Venturi tubes, we can understand the changes

fluids take under this specific circumstance.