Multi-Disciplinary Senior Design Conference Kate Gleason College of Engineering Rochester Institute of Technology Rochester, New York 14623

alligatorsavoryUrban and Civil

Nov 26, 2013 (3 years and 4 months ago)



Design Conference

Kate Gleason College of Engineering

Rochester Institute of Technology


New York


Copyright © 20

Rochester Institute of Technology

Project Number

John Blamer (ME)

Team Leader

Promit Bagchi (ME)

Lead Engineer

Elliot Kendall (ME)

Hydronics Engineer

Matthias Purvis (ME)

Operations Engineer


The objectives of this project revolve around the
installation of a donated industrial process compressor.
After delivery and installation, we will purchase and
implement a data acquisition system, characterize the
compressor’s beginning of life performanc
e, and
complete Vibrations and Thermal
Fluids lab exercises.
These labs will be used by RIT undergraduate classes.
Research on the compressor will be done for Dresser
Rand. Our faculty guide and customer Dr. Jason
Kolodziej of the Mechanical Engineering de
at RIT specified needs which we translated into
engineering specifications. To ensure timely
completion of all the project goals we delegated the
primary responsibilities of mounting, cooling, data
acquisition, safety, and other installation tasks

individual group members. By the end of the project
we successfully designed and implemented mounting,
and cooling systems, installed the compressor,
acquired and installed some components of the data
acquisition system, and revised and tested a vibrat
lab. Each of these systems and procedures is outlined
in detail throughout the paper.


Rand is a multinational corporation
headquartered in

Houston, Texas
. It
provides a wide
range of technology, products, and services used for
g energy and natural resources including
reciprocating compressors.

gas compressor

is a
mechanical device that increases the




by reducing its


The horizontal single
throw reciprocating compressor
is one of the simplest
and most basic of all compressor
designs. These compressors have unmatched
versatility and dependability. Basic components are
engineered, pre
assembled and pre
tested. The
final selection is verified by special computer
programs that will calculate th
e performance, select
cylinders, frames and distance pieces, and compute the

and inertia loads and torsion

stresses which will be
encountered in your service.

Compressor Features:

All running gear components are pressure
with filtered oil,
distributed through inter
nal rifle
drilled passages.
The complete system is protected by
an oil pressure shutdown switch. Units have a pressure
gauge and a crankcase window
type oil level indicator.
Oil filter is automotive cartridge
type for easy

Main, crankpin and crosshead pin bearings are full

free to rotate on the bearing journal and
within the bearing housing. Due to the unique design
having pressurized oil on both sides of the floating
bearings, friction is reduced and wear is

evenly on both sides of the bearing. Pistons are
provided with latest technology, non
metallic piston
ring and rider band materials to ensure maximum

The previous senior design team (P09452) to work on
this compressor project did

a lot of work in
preparation for the installation. The team worked with
Jason Vigil, a Structural PE, to design a solution to
this structural problem. To verify the PE’s result, a
model in ANSYS Finite Element Analysis software
Proceedings of


Design Conference


Project P

was created and simulated.
The delivery path for the
installation of the compressor was evaluated to make
sure that there is enough clearance, and the structure of
floor would not be compromised.

For our senior design our focus was to review and
verify all the previous team’s w
ork and continue
through the final steps to a successful installation and
recording the compressor’s beginning of life


Our main goal for this project is to get the compressor
installed and fully functional. This

includes building a
safe and professional testing area by cleaning and
organizing the compressor room. Our project also
includes drafting a simple but thorough easy
guide that faculty and students can reference for the
entire operating procedures.

Another major aspect of this project is to connect a
DAQ system and collect beginning of life
characteristics of compressor performance and
compare baseline data to provided specifications. This
data is critical for Dr. Kolodziej’s future diagnostic

The third leg of this project is preparing the
compressor for educational use. This

and updating the existing
ibrations and
Fluids labs.

In order to meet our customer needs our team divided
and focused on three major parts:
Mounting the
compressor to the floor in the chosen area, designing
and building a water cooling system, connecting and
calibrating a DAQ system and all necessary sensors.


There were a series of engineering specifications
regarding the coo
ling system which included
providing four gallons of 80°F water, per minute, to
the compressor. The previous team assigned to this
project (P09452) outlined a basic system design
consisting of a circulation pump, heat exchanger,
water storage tank, diverti
ng valve, and immersion
probe heater. The team elected to use a closed loop
heat exchanger utilizing RIT's chilled water supply
instead of running city water directly through the
compressor to "conserve water, and keep with the
sustainability track of the
project." When we took over
the project the heat exchanger had already been
installed in the test cell. Their system design did not
include any specifications or supporting calculations
for the other components of the cooling system so we
decided to undert
ake a full analysis. Determining the
necessary head the circulation pump needed to
overcome revealed extremely high head losses in both
the compressor and the heat exchanger, and made us
aware that we would require a larger pump than
previously expected. R
esearch revealed that a multi
stage pump would be ideal for this kind of application,
where the pump must be able to supply only a small
flow rate at a very high pressure. During the process of
investigating different pump options, we came across a
ration pump assembly constructed by John
Wellin, one of the professors in the Mechanical
Engineering department. The pump rig consisted of a
1/3 hp pump, 50 gallon storage tank, flow meters, and
a flow regulating valve. The system was not in use at
the tim
e and Professor Wellin said we could use some
of the components, or the entire system, for our
project. This system suited our needs very well and
adopting it saved us a lot of work, but it was lacking
several critical components. The most immediate issue
was that the pump was not powerful enough to
produce the required amount of head. Preliminary
calculations revealed that the existing impellor could
be salvaged, and would be capable of producing the
required pressure if it were run at twice the original
peed. Hydraulic power requirements with a
reasonable pump efficiency estimation were used to
determine that a 1/3 hp motor operating at 3450 rpm
would satisfy head and flow requirements. Installing
and testing the motor revealed that it could indeed
e the desired pressure and flow rate, however,
running the pump for longer than 30 minutes caused
the motor to overheat and shut itself down to prevent
It was

concluded that the actual impell
efficiency was drastically lower than expected because

the pump was being used at the low extreme of its
operating range and driven at twice the designed
operating speed. At this point the best option was to
abandon the original pump and purchase a new pump
that was designed to suit our operating requirements
A 1/3 hp, 8 stage, Dayton pump was purchased and
installed, and it produces the desired specifications
without any problems.

Once the pressure and flow conditions were satisfied
the next step was to produce the correct temperature
cooling water. The prev
ious team decided that the best
course of action would be to use a flow diverting valve
with a temperature sensor in the reservoir that would
divert hot water to the heat exchanger when the
temperature in the tank became too high, as shown in
Figure 1. The
y also proposed an immersion heater to
heat the water in the tank to 80°F when the
compressor is started cold.

Proceedings of the

Design Conference


Copyright ©

Rochester Institute of Technology


Preliminary calculations for the required size of the
immersion heater revealed that a large and expensive
heater would be needed. After speaking with Scott
Delmotte, our contact at Dresser
Rand, we concluded
that the heating element would be unnecessary a
cooling the compressor with room temperature water
at start up would not damage it. After researching
diverting valves we decided that a less expensive and
more accurate alternative would be to utilize a
thermostatic mixing valve. This valve produces th
same result as the diverting valve, but it is simpler and
less expensive. The final system design is shown in
Figure 2.


After acquiring and assembling the components the
system was tested without the compressor, and th
pump was capable of generating an adequate flow rate.
The mixing valve can be adjusted to produce
temperatures of 70° to 120° F. The output temperature
was calibrated by running the compressor at steady
state for over 2 hours and monitoring the temperatu
of the water being supplied to the compressor and
adjusting the valve as needed to produce the desired


The D
R compressor requires a very robust and well
engineered mounting system to secure it to the
building structure.

The in
itial concern was the structural integrity of the
floor on which the compressor would be placed.
Compressors of this size are usually mounted on
ground level, on top of dedicated concrete slabs over a
foot thick. Unlike typical installations, we were
ng the compressor in a second story Machine
Shop. The floor here is a 5” thick concrete slab. In
previous years, the original floor was deemed
inadequate to support the over 7,000 lb compressor
with the help of Structural Engineer Jason Vigil. SD
team P094
52 oversaw the installation of two I
under the floor to add support in preparation for the
compressor arrival.

Due to its horizontal orientation, mass, and low
operating speed of 360 rpm, the compressor can
introduce a significant resonating frequen
cy to the
building. In typical installations the concrete
foundation would be capable of absorbing such
frequencies. The previous Senior Design team did not
utilize any type of vibration isolators, and proposed
the unit be bolted directly to the floor. Aft
er consulting
with several Subject Matter Experts, we determined
this would not be adequate. If bolted directly to the
floor, the vibrations produced from operation would
most probably be felt elsewhere in the building. We
were forced to look at other solu
tions to prevent
transmitting vibrations to the surrounding Machine
Shop. Lord Corporation did a vibration analysis for
our application and recommended the use of their
rubber isolation mounts. They were kind enough to
donate 10 of these mounts for our pro
ject. These
mounts are capable of supporting 1000 lbs each, and
should absorb the vibration of the reciprocating

The vibration mounts had bolt holes along their
longitudinal axis. When bolting them to the I
frame of the compressor, the bo
lt holes would interfere
with the central I
beam web. Several concepts were
considered for attaching the adapter plate and mount to
the compressor. These concepts ranged from welding
the adapter plate to the compressor frame, to drilling
new holes in the
frame to accept bolts. Welding was
eliminated due to the risk of melting the rubber mount
material, and the potential difficulty if the compressor
needed to be moved in the future. The compressor
frame was pre
drilled for 10 mounting bolts along the
ter. These holes were only on the outside of the
beam web. If only these outer holes were used to
attach the frame to the mounts, we would not have an
even, positive attachment. This could hinder the
effectiveness of the mounts. In order to bolt the fram
to the mounts on both sides of the I
beam web, we had
to drill holes on the inner flange. This drilling was
only possible from the bottom, and required the unit to
be suspended in the air. Boulter Rigging had the
equipment necessary for this type of task
, and
completed the drilling for us. The new holes mirrored
the original 10 mounting holes, and gave us a positive
attachment point for each vibration mount.

Ultimately, we designed and fabricated adapter plates
out of ¾” steel plate. These plates bolted
to the
Proceedings of


Design Conference


Project P

vibration mounts through counter
bored holes, and to
the compressor frame through two threaded holes
which straddled the I
beam web.

Figure 3

The next point of concern was the attachment of the
vibration mount to the cement floor. This
point was crucial, and needed to be robust for safety
and to minimize vibration. The original layout
recommended by the Structural Engineer was to drill
and bolt through the floor in 10 locations around the
frame. Since we modified the attachmen
t method to
include the vibration mounts, we knew this would
change. In addition to bolting through the floor, we
considered using concrete anchors which did not
penetrate all the way through the floor, and an epoxy
grout to adhere the mounts to the floor.

Each mount would require at least two bolts securing
it to the floor. Each vibration mount had four
mounting holes on the bottom plate. If we utilized
each of those holes we would have 40 holes through
the floor. We deemed that many holes and bolts to be
unnecessary. The structural integrity of the floor after
40 holes were drilled was a primary concern. After
consulting with the Structural Engineer, we decided to
install two bolts per mount (opposing corners) which
made for 20 holes drilled through the fl

Figure 4

A CAD model of the compressor frame, concrete
floor, mounting system and potentially interfering
structures was constructed. A large concrete beam
was then found directly under the bolting locations for
one side of the compressor. If the compressor was
stalled in this original location we would not have
been able to drill through the large concrete beams. To
resolve this issue the compressor location was moved
12 inches over towards the center of the room. This
deviation actually became an improvement. W
ith the
original location, access to the control panel would
have been limited. In the new location, there is
sufficient room around the perimeter of the unit.


The installation plan we formulated included securing
the compressor to the fl
oor, connecting the cooling
system, and working with Dresser
Rand to install
several items which included belts, cylinder packing,
oil scraping rings, and other miscellaneous parts.
These tasks were completed, but not without several
hurdles. The first pr
oblem, which arose the day before
the compressor was scheduled to arrive, was the
doorway being about 3 inches too narrow for the
compressor to fit through. This width was
unaccounted for and required immediate attention.
Once the problem was identified, p
roblem solving
techniques were utilized. The compressor diagrams
and pictures taken of the compressor at Boulter
Rigging, the company delivering the unit, were
reviewed. After speaking with Boulter and analyzing
the situation, several solutions were sugges
ted. The
most practical solutions included moving the
compressor into the room at an angle to gain
clearance, widening the door frame to accommodate
the extra width, or removing the control panel and
placing it in a location that would not interfere with
he installation. After analyzing the possible solutions
it was decided that the best course of action would be
to cut off the control panel and move it several inches
board. Boulter Rigging was capable of such
modifications. On the morning of scheduled

the Senior Design team traveled to the Boulter facility
to finalize the plan. The control panel was cut off, and
relocated about 3.5 inches and re
welded to the frame.
Boulter Rigging did a superb job in handling such a
major task, last minute.

The modification was successful and the compressor
was delivered on schedule, but another setback was
encountered upon installation. Drilling bolt holes in
the test cell floor to secure the compressor caused
3 inch

diameter cones of concrete

break free fr
om the underside of the floor. Rather than
continuing to damage the floor, the use of concrete
anchors was substituted for bolts through the floor.
This eliminated the need to drill through the floor and
create more damage. Seven of the 20

mounting points
were attached with bolts, and the remaining 13 were
secured with anchors. The ½” anchors only required a
hole about 2 inches into the floor. They expanded and
provided over 2500 lbs of anchoring force.

The resulting damage

threatened the i
ntegrity of the
floor structure, and inhibited the effectiveness of the
3”x3” steel washers.
The possible solutions that we

devised for this

were to fabricate larger washers,
use ten foot steel plates that would run the length of
the compressor on
each side and secure all of the bolts,
and fill the holes with cement and use larger, 6”x6”,
Proceedings of the

Design Conference


Copyright ©

Rochester Institute of Technology

washer plates.
The chosen solution was to

fill the
holes with cement and use the larger

es. This
solution was quick,
simple and provided


The repair was approved by Structural
Engineer, Jason Vigil of Jensen Engineering.

Once the compressor was bolted

sheave was installed

and belts


This was a
guess and check process which took quite a long time
of the tasks were completed
. Another task
that was


in the installation


attaching the cylinder packing and oil scrapping rings.
he manual specified

which type of cylinder packing
or oil scrapping ring, how many of each and the
on in which they should be placed on the
connecting rod. The cylinder packing and oil
scrapping rings seal the crankcase from oil leaks.

When thes
e tasks were completed the compressor was
turned on
and it had no mechanical i
When the
compressor was
initially turned on,
it oscillated more
than expected
It moved

enough to make the vibration

by about 5/8 of an inch
, but the
isolators successfully prevented any noticeable
vibration from being transmitted to the floor
Therefore, the mounts

donated by LORD Corp.
successfully dampened most, if not all of the input
force from the compressor to the ground.


The compressor has been successfully installed and
prepared for normal operation. The cooling system
meets all requ
ired specifications. The adaptor plates
which were fabricated for the LORD Corporation
lattice mounts work as intended. The isolation
dampers prevent vibration transmission to the
building. The DAQ system was set
up, tested and
The compressor’s
vibration was
characterized by u
sing Lab

to gather data and

acceleration vs. t
ime plot along with a
to show the compressors motion in
each axis as well as its operating frequency.


Unfortunately, the full scope of this project was not
en through. Due to unforeseen
circumstances, some setbacks were encountered such
as the modifications required to fit the unit through the
test cell door. Such errors should have been caugh
t in
advance, but it would have been impossible to account
for every possible setback. Although these problems
did introduce schedule deviations, they were dealt with
in a timely and professional matter. For example, the
control panel modification was comp
leted on the day
of delivery, but the compressor was still delivered on

Some setbacks were encountered due to errors in the
work of previous groups. Although the work was done
very well, it should have been double checked and
verified prior to the compressor installation.

Work on the educational labs was not completed as
ed. This section of the project will have to be
completed by either future Senior Design groups or
students. Further work should be done on the DAQ
system to expand its capabilities and collect broader

The cooling system should be rebuilt with a st
frame construction. A smaller, more suitable water
reservoir should be utilized. This would decrease the
thermal mass of the cooling system, allowing for faster
up times. The steel frame design can be made to
better accommodate the cooling system
components in
a more compact configuration. It would also be more
durable than the current wood construction.

When possible, the adjacent room to the test cell
should be utilized as a control room. This room is
currently occupied by the RIT Formula SAE Te
am. By
moving the DAQ hardware into this room, it will
create a more suitable environment for operators to
collect data and analyze it on computers. Under load,
the compressor can become relatively loud with the
discharge air exiting within the room. If op
erators are
within the room while the compressor is at full load,
ear protection is recommended.

The control panel must be reconfigured to accept the
correct switch
type for the power on. Currently a
momentary switch is in place while the relay
is meant
for a latch
style. This switch should be already ordered
and in
house and just needs to be installed. An
electrician will be needed to rewire the cooling system
motor outlet back to standard 110V. The motor will
also need to be reconfigured for this voltag

The orifice tank, where the compressed air is
discharged can potentially be placed in the basement if
approved by RIT and Dave Hathaway. This will
reduce the noise in the test cell. There is plenty of air
line to run the tank in the basement. The Rose
pressure transducer can also be mounted in the
basement, with the signal wires run up to the DAQ

Proceedings of


Design Conference


Project P


Senior Design team P11452 would like to sincerely
thank the following groups and organizations for their
generous contribut
ions to our project.


Dr. Jason Kolodziej



Bill Nowak



Scott Delmotte

Dresser Rand


Dresser Rand Corporation


LORD Corporation


John Wellin



Dave Hathaway & ME Machine Shop Staff