Ram Pump Fabrication Manual
RamPumpFabricationManual
User Manual: Pdf
Open the PDF directly: View PDF 
.
Page Count: 8
| Download | |
| Open PDF In Browser | View PDF |
Ram Pump Operation and Testing Manual
Spring 2018
Maile McCann, Will Lopez, and Steven Lopez
The goal of the Ram Pump subteam is to optimize the pumping efficiency of the hydraulic ram pump, where efficiency is measured in energy of water entering the pump over the energy of the
water pumped. The hydraulic ram pump pumps clean water through the plumbing system of an AguaClara water treatment plant, which provides treated water for use in the chemical dosing
system as well as sinks and toilets, and saves operators time and energy transporting treated water up by hand. The ram pump system is entirely electricity-free.
A major goal of Spring 2018 was to examine maximum energy efficiency. The value of calculating experimental energy efficiency is that the team is then able to compare the values to calculated
theoretical energy efficiency. From there, we can exclude terms in the theoretical calculations to determine the main contributors to inefficiency.
This manual outlines procurement, fabrication, testing, and cleaning of the current hydraulic ram pump model as of Spring 2018.
Background and Fabrication Details
The ram pump itself consists of tw o 1.5" threaded metal check valves connected by a custom PVC fitting. The upper check valve contains the functional interior of the ram pump. The interior of the
ram pump consists of a threaded metal rod w ith tw o metal standoffs screw ed on, and a hole in one end that has been tapped to accommodate a threaded metal plate. The spring fits into the interior
by fitting concentrically over the end of the metal rod below the standoff. The purpose of the standoffs is to determine the maximum oscillatory period of the pump.
The threaded standoffs can screw up or dow n to accommodate a larger or smaller distance of oscillation. The metal plate is able to screw into the stop of the metal rod. The rod drives the
oscillation of the metal plate through the check valve via the spring force. The metal rod is able to slip inside the bottom check valve, holding it in place horizontally but allow ing vertical oscillation.
The spring, w ith a larger radius than the rod, does not fit through the bottom check valve opening, and instead compresses w hen, on a dow nw ard drive, the metal drive rod encounters the bottom
check valve. The compression of the spring causes an upw ard spring force that subsequently drives the metal plate back upw ard. See Figure 2 for interior setup in the inverted configuration.
The driving force of the ram pump is provided not by the force of the plate, how ever, but the pressure gradient created by the oscillation of the upper check valve. When the plate slams shut, the
pressure above the upper check valve exceeds the pressure external to the effluent tube, thus driving flow through the effluent. When pressure has been released, the spring is able to overcome
the sole hydrostatic force on the metal plate, and thus the plate oscillates open. When the hydrodynamic force of the drive flow hits a maximum, the drag on the plate closes the plate again,
reinitiating the pressure gradient.
The upper check valve connects to the PVC drive pipe w hich provides the influent flow , or the treated w ater exiting the AguaClara plant. The connection is made by a PVC male to male connector
that screw s into a male metal nipple and into the top check valve. For lab testing purposes, previous subteams have established system to simulate flow from the plant. A sump pump carries w ater
to a raised tank, and the w ater from this tank flow s the drive pipe w hich initiates the pumping process. The elevation of the head tank creates the energy input in order to drive the pressure
oscillation in the ram pump. See Figure 1 for exterior setup.
Figure 1: Exterior Ram Pum p Set Up. This is the schem atic of the ram pum p fittings from an exterior view, show ing how each piece threads to the next.
Calculating Minimum Spring Force
The compressed spring w hen the plate is in the closed position causes an upw ard force to open the plate, proportional to the k constant of the spring. In order to increase the time in the closed
position and thus increase w ater pumped (see Python Calculations section for mathematical explanation) the k value for the spring should be minimized.
The spring force at most overcome the hydrostatic force w hen the plate is closed. Therefore, to find the minimum possible spring force, the team created a force balance in a python code to find
the k value that w ill produce enough force to overcome the hydrostatic force of the w ater in the drive pipe.
# This code can be used to determine the minimum spring force necessary
# Force must be greater than hydrostatic force in order to reopen
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from aide_design import physchem as pc
from aide_design.units import unit_registry as u
from aide_design import utility as ut
#
#
#
#
#
h_drive
A_drive
x_max =
h_eff =
A_eff =
= height of drive pipe
= Area of drive pipe
maximum spring compression distance, equal to plate ampltitude
Necessary pumping height
Area of effluent flow
# Minimum force balance
# k_min = minimum k value needed to open from closed position
temp = u.Quantity(22, u.degC)
# Assume room temperature.
rho = pc.density_water(temp)
g = pc.gravity
x_max = .045*u.m
h_drive = .51*u.m
d_drive = 0.025273*u.m
A_drive = pc.area_circle(d_drive)
P_hydrostatic = rho*g*h_drive
k_min = P_hydrostatic*A_drive/x_max
For the Spring 2018 semester, the team calculated a minimum k constant of .38. The team then ordered springs of that k constant at varying lengths to look for a correlation betw een length and
pumping efficiency.
Check Valve Inverted Configuration
A possible modification to the current set up to increase efficiency w as the inverted configuration. The purpose of the inverted check valve configuration is to reverse the force balance in the ram
pump. The configuration involves reversing the upper valve so that the upw ard spring force pushes the plate into the closed position, and the hydrostatic force of w ater opens the valve. The
effect is that the spring force has to only resist the hydrostatic force in the closed position, thus potentially lengthening the time in the closed position. Due to the increase in time for the closed
position, more w ater flow s through the effluent therein increasing the efficiency.
The CEE machine shop is equipped to fabricate the inverted valve. The inverted check valve consists of a standard metal check valve w ith the follow ing modifications. Contact Tim or Paul of the
CEE Machine Shop (in the basement of Hollister next to the AguaClara lab) to assist w ith the modifications to the unaltered metal check valve. They are both equipped to operate the drill and tap for
this modification. Additionally, Tim or Paul are both resources to assist in follow -up alterations or fabrication questions.
Check Valve Modification
1. Drill a .25" diameter hole 2.4" up from the bottom of the valve and another .5" hole at the same elevation at the opposite side of the valve, as visually represented in Figure 2: Interior Ram Pump
Set Up below .
2. Tap both holes in order to accommodate threaded pipe fittings for the pressure test valve and the effluent valve, as show n in Figure 2.
3. Screw in the pressure test valve and the effluent valve.
4. Apply acrylic flex sealant tape around connection betw een threads and valve in order to ensure w ater tightness.
Figure 2: Interior Ram Pum p Set Up After Modification. From this im age, you can see how the m etal plate fits into the upper check valve and attaches to the m etal rod. The
spring, as show n, inserts concentrically around the drive rod and it held in place by the bottom check valve.
PVC Fitting Fabrication
This component connects the tw o metal check valves to hold the inner metal rod in place. This fitting w as fabricated by this semester's subteam and can be reused until w ater tightness fails or the
fitting undergoes significant internal pitting, how ever this section w ill outline how to fabricate a new piece.
1. Purchase tw o 1" male connectors for w hich one side is threaded and the other is smooth.
2. Depending on the length of the rod and spring on interior of pump, use band saw to cut one connector so that both connectors together allow the inner rod to rest on spring inside valves
w ithout compression initially. The spring w ill be approximately 2.5" in length, but make measurements and mark male connectors to appropriate size for current configuration being tested.
3. Weld the smooth ends together using PVC w elder.
4. Allow the apparatus to cool completely before attaching to ram pump. This is crucial to ensure that the w eld sets correctly and the piece doesn't break during testing.
Figure 3: Spring 2018 Fabricated Male to Male PVC fitting. This is the current team 's fitting fabricated in Spring 2018, w here you can see the w elds done w ith the hand held
PVC w elder.
Special Components
The ram pump consists of multiple components that are not easily accessible in the AguaClara research lab or on campus. These pieces must be specially ordered, and are listed below .
1. Check valve
Team can order new check valves from McMaster Carr online store.
Generally, check valves can be used from the previous year unless severely deformed or threading has degraded.
https://w w w .mcmaster.com/#check-valves/=1chzv9q
2. Metal plates
The metal plates can be reused from previous years.
How ever, if new plates need to be ordered they are generally only available w ith the purchase of a new check valve.
https://w w w .mcmaster.com/#check-valves/=1chzv9q
3. Springs
The team purchased spring of desired k constants from Century Spring Corp, but most testing can be accomplished w ith springs purchased in previous semesters.
If springs must be ordered, contact grader that is in charge of ordering parts w ith a copy of your shopping cart and instructions on how to order. For Spring 2018, this w as Casey Ching.
https://w w w .centuryspring.com/?
matchtype=e&netw ork=g&device=c&adposition=1t1&keyw ord=century%20spring%20corp&gclid=EAIaIQobChMIgYee1ejJ2gIVhFmGCh0IGgQZEAAYASAAEgLUXPD_Bw E
i. Metal rod and standoffs
The unit consists of an 1/4 inch threaded metal rod and tw o metal standoffs that thread onto the rod.
Generally, this piece does not have to be replaced unless lost. The rod does not experience excessive w ear.
Experimental Methods
The follow ing instruction details how to conduct energy efficiency testing of the ram pump. This determines how efficiently the system pumps effluent for the amount of volumetric flow entering the
pump from of the drive pipe. Energy of both the effluent pumped as w ell as the incoming drive flow is the product of flow rate and hydraulic head. The energy pumped is the product of volumetric
effluent flow and effluent head, or the equivalent pressure that the ram pump is pumping against to pump the w ater to a higher elevation. The pressure head can be measured using a 7 kPa
pressure sensor linked to ProCoDA. Volumetric flow rate is determined by recording the amount of time it takes to pump a certain volume, as show n in the follow ing equation:
$Q= \frac{V}{t}$
The overall equation for energy efficiency of the ram pump is as follow s:
###### Figure 5: Energy Efficiency Equation, w here
###### $Q_{eff}$ = volumetric flow rate of effluent
###### $h_{eff}$ = pressure head of effluent being pumped, eg: how high the effluent is theoretically being pumped
###### $Q_{w aste}$ = volumetric flow rate of the w asted w ater during the same time interval measured for the effluent flow rate
###### $h_{drive}$ = height of the drive pipe, representative of the incoming energy to the ram pump
Energy efficiency w ill become very useful in the final phase of Spring 2018's semester goals. The value of calculating experimental energy efficiency is that the team is then able to compare the
values to calculated theoretical energy efficiency. From there, w e can exclude terms in the theoretical calculations to determine the main contributors to inefficiency. For example, w e can exclude
the minor headloss term in the theotrical efficiency calculation to see how much minor headloss reduces effiency and if this is an issue to address as w ell as spring force. This analysis w ill be the
focus of the last third of the semester, and w ill be further detailed in the final manual.
Experiment
Figure 4: Standard Configuration: the setup of the apparatus that the team uses in the laboratory to conduct all experim entation
1. Assemble ram pump in the configuration show n in the figure above, w ith the air chamber connected from the effluent pipe w ith a push to connect T fitting.
2. Fill the w aste w ater bucket roughly half w ay to the brim.
3. Attach a push-to-connect ball valve to the end of the effluent flow pipe in order to create pressure w ithin the air chamber.
4. Connect a pressure sensor to the effluent outlet to measure the effluent pressure of the system using ProCoDA
5. Open the bucket connector valve and close the drive pipe valve.
6. Plug in the electric sum pump and allow it to run until most of the w ater is removed from the collection bucket. Then close the bucket connector valve and unplug the pump. Mark the w ater level
in the collection bucket using tape.
7. Connect the air chamber to the effluent pipe as displayed in Figure 4. Attach a 200 kPa pressure sensor to the top of the air chamber and start ProCoDA data collection. Open the drive pipe
valve w hile starting a stopw atch at the same time.
Run the ram pump until roughly 30 seconds elapse. Stop the timer and close the drive pipe valve simultaneously. Stop the ProCoDA data collection.
8. Measure the volume of w ater in collection bucket underneath the drive pipe by subtracting the final volume by the initial volume marked by the tape. Divide this value by the time measured to
compute the flow rate of the effluent and w aste streams. The w aste stream is the flow that comes out of the bottom of the system and is recycled back through the sum pump. The effluent is
the flow that that leaves through the check valve and is pumped to a designated height.
9. Measure the length of the drive pipe and compute the effluent head by dividing the pressure measured from the sensor using ProCoDA by the density of w ater multiplied by the gravitational
acceleration constant.
$h = \frac{P}{\rho * g}$
9. Calculate energy efficiency using the pressure data from ProCoDa. This w ill be done by using the initial volume of the chamber and Boyle's Law as listed below . The initial volume and the
pressure at tw o points are used to calculate the new volume for the second pressure. This is continuously interpolated until volume is obtained for every pressure measurement obtained.
Using these volumes and the time w e can calculate a volumetric flow rate.
$ P_{1} * V_{1} = P_{2}* V_{2} $
Writing the Python code for this data analysis is part of the team's focus for the last third of the semester, and w ill be included in the final research report.
Cleaning Procedure
1. Run DI Water through the system
2. Close the valve to the sum pump w hile the w ater is running through to accumulate the w ater in the w aste bucket
3. Drain the w aste valve and remove all attached tubing
4. Run a 9:1 w ater to bleach solution through the tubing using a peristaltic pump
5. Allow the ram pump and PVC to air dry over night
6. Make sure there are no spills around the ram pump or on the lab station
7. Remove excess and unnecessary clutter from the lab station
Experimental Checklist
1. Check that the w ater level in the w aste w ater is above the outlet to the sum pump
2. Make sure every valve is set to open
3. Check that all the fittings are secured to prevent w ater leaks
4. Ensure that the spring and the rod are w ell aligned inside the valve
5. Make sure there is no w ater on the surface of the table w hen plugging in the sum pump
6. Check that the effluent flow valve is discharging into the container
7. Check that there are no obstructions in the tubing connected to the pressure sensor
ProCoDA Methods
ProCoDA can be used to measure the pressure in the effluent valve in order to determine the head of the effluent valve w hen efficiency testing. In order to determine the max height that the
effluent w ould achieve, measure the pressure and use equation
$P=\rho * g * h$
and rearranging to obtain
$h =\frac{P}{\rho*g}$
This height is then plotted using every pressure obtained to determine how height varies w ith pressure. To accomplish this, attach a short (no longer than 12 in) 1/4 inch plastic flex tube into the
pressure test valve. Plug the pressure sensor into a port on the PC w here ProCoDA w ill be accessed. Run the pump as explained in experimental procedures, this time recording pressure
readings. After 30 seconds, stop the data file and examine the points of highest pressure. This w ill be the pressure w hen the plate just opened and effluent w as being pumped. Average the
highest pressure values in the pressure oscillation to obtain your average effluent pressure.
Python Calculations
As previously discussed, a goal of Spring 2018's subteam w as to find a maximum efficiency value for the current apparatus in order to determine causes of inefficiency. To determine maximum
efficiency, the team w rote a Python code that can be transferred to future teams in order to have a max efficiency value each semester for the current apparatus.
The governing equations for w ater pumped in both the deceleration and acceleration phases are the integrals of velocity in both cases.
While the plate is open, the velocity is accelerating and is governed by the follow ing equation:
Figure 5: Full equation for velocity of a hydraulic transient during the acceleration phase
In order to determine headloss coefficients experimentally, the team measured the terminal velocity of w ater flow ing through the drive pipe and used the follow ing relationship to back calculate the
headloss term:
Figure 6: Equation for verm inal velocity, from w hich the team extracted a constant for head loss
After the plate closes, w ater begins to decelerate and the velocity during deceleration is governed by F=ma, w ith the force on the w ater caused by the pressure gradient from the drive pipe to the
effluent. The full equation can be found in the Python code at the bottom of this manual.
In order to determine energy, the code takes the integral of each part of the equation. The bound of integration for acceleration are from 0 to the time w hen the plate closes. The bounds of
integration for deceleration are from the time the plate closes to the time velocity goes to 0 (x intercept of deceleration). This is represented graphically below :
Figure 7: Graph produced by Python Code of ideal velocity cycle; before intersection is w hen the w ater is accelerating, after the intersection is deceleration
The final step in calculating efficiency is to multiply the integral of acceleration by the height of the drive pipe to obtain energy input; then multiply deceleration by the height of the effluent by the
effluent pipe to obtain energy pumped. The full code is included below .
Due to lack of time in the semester, the team w as unable to find an accurate experimental value for efficiency. The team performed multiple efficiency tests on the various springs w ith calculated
spring forces. How ever, due to noise in pressure sensor readings the efficiency values w ere extremely inaccurate, averaging almost 60% efficient. See the follow ing github issues for testing
documentation:
Efficiency Testing 4/10
Efficiency Testing 4/11
Efficiency Testing 4/13
Code for Efficiency Calculations
# Acceleration
# L = length of major losses, h_drive
# v_f = velocity of flow when plate is completely open
# k = constant for major and minor loss coefficients
import scipy.integrate as integrate
from aide_design.play import *
t = np.linspace(0, .2, 500) # 500 seconds, 4 intervals
H = 5 # Height of drive pipe.
L = 12 # Length of effluent pipe.
g = pc.gravity.magnitude
v_f = .349 # Experimentally measured terminal velocity.
k = (2*g*H)/(L*v_f ** 2)
def velocity_a(time):
return np.sqrt(2*g*H/k)*np.tanh(time*np.sqrt(g*H*k/(2*L)))
v = velocity_a(t)
plt.plot(t, v)
plt.xlabel('Time')
plt.ylabel('Velocity')
plt.title('Ideal Velocity vs Time')
plt.show()
# Deceleration
# Governing equation is F = ma
#
#
#
#
a
The deceleration of the water after the valve closes is directly proportional
to the corresponding increase in pressure.
The pressure spike generated the moment the valve closes will be measured in
the lab.
= DeltaP/(rho*H)
# Time when plate closes (measured or calculated)
t_c = 0.125
# Velocity at time plate closes.
b = velocity_a(t_c)
# Velocity while plate is closed and water is deccelerating.
def velocity_d(t):
return a*t - t_c*a + b
v_d = velocity_d(t)
# Find t at v=0
b = len(v_d)
for x in range(0, b):
if (v_d[x-1] >= 0 and v_d[x+1] <= 0):
x_intercept = t[x] # Where the deceleration crosses the x axis.
x_intercept
# Plotting both
plt.plot(t, v)
plt.plot(t,v_d)
plt.xlabel('Time')
plt.ylabel('Velocity')
plt.title('Ideal Velocity vs Time')
plt.ylim((0,1))
plt.xlim((0,.14))
plt.show()
# Integrate acceleration and decceleration.
w_int = integrate.quad(velocity_a, 0, t_c)
p_int = integrate.quad(velocity_d, t_c, x_intercept)
p_int
waste = w_int[0]
pumped = p_int[0]
# Calculate efficiency.
h_eff = 10
max_E = (pumped*L)/(waste*H)
max_E * 100
python
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.4 Linearized : No Title : Creator : Producer : Qt 5.5.1 Create Date : 2018:05:18 13:34:52 Page Count : 8EXIF Metadata provided by EXIF.tools