CASE HISTORY#1

BAD COOLING TOWER GEARBOX?

Barry T. Cease
Cease Industrial Consulting
[email protected]
(843) 200-9705

1
 CASE HISTORY – BAD CT GEARBOX?
• The customer reported a terrible noise coming from at least one of
his cooling tower fans (CT fan C). He and his staff were convinced the
gearbox was bad.
• The customer had already prepared to have the gearbox changed, but
wanted confirmation via vibration analysis prior to performing the
work.
• The facility had three identical cooling towers on-site (units A, B & C).
• All three cooling tower fans were 6-blade, driven via right angle
gearboxes (5.5:1 ratios) thru a long spacer shaft by a 4-pole motor
operating on a variable frequency drive.

2
 ON-SITE INSPECTIONS
Before any vibration data was collected on the fans, an on-site inspection of all three
cooling tower fans found the following:
1) The pitch of all blades on all fans were checked for consistency via a scribe mark on
the shroud. The pitch of some blades at both B & C fans were found off by a
significant amount. These blades were adjusted accordingly to conform to the others.
2) The distance from the blade tips to the nearby shroud was checked on all fans and no
significant problems were found.
3) The condition of the jackshaft couplings was inspected and no significant cracking or
swelling of the rubber grommets was noticed at any of the three fans.
4) All jackshafts appeared ok and no obvious distortion or other shaft damage was
noticed.
5) All bolts holding the gearbox to the tower frame appeared tight.
6) Minor oil leaks were noticed at the input shaft seal of each gearbox. This leaking may
have been due to overfilling of oil in the units in the past?
7) The rubber oil drain hoses at each gearbox were discolored, beginning to crack at
spots and generally in bad shape.
3
 VIBRATION DATA
• After the on-site inspections & minor repairs were complete, vibration
measurements were taken on all three cooling tower fans using
magnetically mounted accelerometers on both the motor and gearbox in
the horizontal direction.
• A laser tach was used to measure machine speed and vibration data was
collected with the fans running at both full speed and thru their normal
speed range (30-100% full speed).
• A spectral range of 120-120,000 cpm was used.
• As reported by plant personnel, during vibration testing, an unusual noise
was heard at fan C that wasn’t heard at the other two units. This noise was
evident at the higher speeds only.
• A comparison of vibration data collected from the three fans is shown
below.

4
COOLING TOWER
OVERALL VIBRATION
LEVELS (IPS-PK)

• All gearbox levels were < 0.2

ips-pk (not that bad).
• Gearbox levels at fan “C”
were in-line with the other
two units.
• Motor levels at fan “C” were
clearly higher than the
others (potential problem).

5
COOLING TOWER
VIBRATION
WAVEFORM LEVELS
(G’S-PK-PK)

• All waveform levels were

< 3 g’s-pk-pk (not that bad
for gearboxes).
• Gearbox levels at fan “C”
were the lowest of the
group at ~ 1.2 g’s-pk-pk.
• No significant motor or
gearbox bearing or gear
faults were expected.

6
 VIBRATION SPECTRAL DATA
• Spectral vibration data was collected from the motor & gearbox of each CT fan
and analyzed. This data showed three primary frequencies of vibration at each
machine as follows:
1) Motor speed (1x rpm) at approximately 1,788 cpm.
2) Fan blade-pass frequency at approximately 1,950 cpm.
3) Fan speed (1x rpm) at approximately 325 cpm.
• These CT fans had a greater separation between motor speed & blade-pass
frequency than is normally experienced by the author (ie: 1,950 – 1,788 cpm =
162 cpm or ~ 9% separation) resulting in less beating vibration than normally
experienced by CT fans.
• As a side note, the author remains puzzled as to why many CT fans are designed
with < 5% separation between motor speed & blade-pass frequency. Due to
beating, this small separation of dynamic forces tends to increase vibration levels
at the machine and thus reduce its long-term reliability.

7
CT MOTOR SPECTRAL DATA (A, B & C)
• Vibration levels were clearly
highest at CT fan “C” motor.
• Dominant vibration at CT fan
“C” motor occurred at the
motor speed of ~ 1,778 rpm
(0.24 ips-pk).
• Potential causes for the higher
levels at motor 1x rpm for “C”
fan were unbalance at
coupling or shaft,
misalignment or coupling
problems.
• Vibration at blade-pass
frequency was low at all
motors, but a 2x & 3x blade-
pass vibration component
were noted at “C” motor.

8
CT GEARBOX SPECTRAL DATA • Vibration levels were only slightly
higher at CT fan “C” gearbox.
• Dominant vibration at CT fan “C”
gearbox occurred at 1x & 2x blade-
pass frequency as well as motor
speed.
• Vibration levels at 1x blade-pass
frequency were much lower at fans
“A” & “B” and no significant vibration
at 2x or 3x blade-pass were noted at
these two fans.
• It is normal to see some level of 1x
blade-pass at CT fans, but abnormal
to see 2x blade-pass levels anywhere
near those at 1x blade-pass
frequency (signals a problem).
• Common causes of high vibration at
blade-pass frequency are improper
blade pitch, unequal flow, problems
with the fan shroud, or resonance.

9
CT GEARBOX SPECTRAL DATA (DISPLACEMENT)
• By converting our
spectral units from
velocity to displacement,
a minor fan balance issue
at CT fan “B” becomes
clear (5 mills-pk-pk).

10
GEARBOX SPECTRAL & WAVEFORM DATA
• From the waveform
vibration data below,
note how beating
vibration is small at CT
fan “C” despite the fact
that it had the highest
levels of vibration at 1x &
2x blade-pass frequency.
• It is the opinion of the
author that CT fans with
a 10% or greater
separation between
motor speed & 1x blade-
pass frequency would
represent one aspect of
good design for these
machines.

11
RUNNING INSPECTION OF CT FAN “C”
• After all vibration data was collected, an on-site running inspection of CT fan “C” was performed in an
effort to help determine the source of the unusual noise at this fan during operation.
• This running inspection identified a portion of the shroud that was cracked and only visible at certain
fan speeds when it apparently resonated and made a terrible noise.

12
RUNNING INSPECTION OF CT FAN “C”
• The running inspection found a cracked portion of the fan shroud that was clearly visible from above at
the higher fan speeds when it began vibrating excessively coming close to touching the fan blades – this
no doubt was the source of the terrible noise heard at this fan.

13
CONCLUSIONS & RECOMMENDATIONS (GENERAL)
1) All three cooling tower fans could benefit from replacing their rubber oil
drain hoses with SS braided types that won’t deteriorate overtime and
potentially cause a failure of the gearbox due to loss of oil.
2) No obvious damage or deterioration of the rubber grommets at the
couplings of the jackshafts were noted during this inspection, however,
from past experience, it might be a good idea to have these inexpensive
grommets replaced every other year or so.
3) Lubrication is the lifeblood of rotating equipment. To achieve long life,
please ensure that the proper type & quantity of oil is being used in all
gearboxes. In my opinion, the oil in both units should be changed every
two years or in compliance with OEM guidelines.
14
 CONCLUSIONS & RECOMMENDATIONS (CT-A & CT-B)

• Cooling Tower Fan A: On-site inspections as well as the vibration data

collected showed no signs of significant problems with this fan at this
time. Other than the general comments mentioned above for all
fans, no specific recommendations are made at this time.
• Cooling Tower Fan B: This fan suffers from a higher level of unbalance
when compared to the others (ie: compare fan A (0.8 mills), fan B (5
mills), and fan C (1.1 mills). Having this fan balanced at your next
convenient opportunity would reduce its vibration levels and likely
extend its life.

15
 CONCLUSIONS & RECOMMENDATIONS (CT-C)
1) The highest levels of vibration at both the motor speed of 1,788 cpm and the
blade-pass frequency of 1,940 cpm were measured at fan “C”.
2) As noted earlier, an unusual noise was clearly heard at this fan during operation
that simply wasn’t heard at the others. Knowing of the high levels at blade-pass
frequency and hearing the noise, I decided to do a running inspection of the fan.
3) From this inspection, I discovered a portion of the fiberglass shroud cracked and
the steel grating covering the top of the fan loose at spots.
4) Excessive vibration, noise and visible movement of the shroud itself in the general
area of the crack were observed while fan “C” was running at the higher speeds.
Similar problems were simply not found at any other part of this fan’s shroud or at
the shrouds of any of the other fans.
5) In my view, the high vibration at blade-pass frequency at this fan is in large part
due to this damaged shroud. Please have this shroud & grating repaired or
replaced as needed, and check all bolts holding the shroud to the structure to
ensure they are all tight.
6) Consider an alignment check at this fan at your next opportunity.
16