Purdue University
Purdue e-Pubs
International Compressor Engineering Conference
School of Mechanical Engineering
1980
Investigation of Non-Lubricated Piston Ring
Problems - Cause, Effect and Solution
R.F. M. Tremain
Follow this and additional works at: http://docs.lib.purdue.edu/icec
Tremain, R.F. M., "Investigation of Non-Lubricated Piston Ring Problems - Cause, Effect and Solution" (1980). International
Compressor Engineering Conference. Paper 307.
http://docs.lib.purdue.edu/icec/307
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] for
additional information.
Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W. Herrick Laboratories at https://engineering.purdue.edu/
Herrick/Events/orderlit.html
INVESTIGATION OF NON-LUBRICATED PISTON RING PROBLENS - CAUSE, EFFECT AND SOLUTION
Hoger F. M. Tremain
Nobrac Carbon Ltd., Lancing, England.
SYNOPSIS
Many ring problems in petrochemical and process
gas industries that have occurred over the last
decade on reciprocating non-lubricated compressors
are examined.
The causes, effects and solutions are correlated
by analysis of the methods by which the problems
were overcome in applied situations, a few of
which are examined in depth for illustrative
purposes.
Conclusions are drawn as to good current practice
for the achievement of reliable piston ring
operation in non-lubricated reciprocating process
gas compressors.
INTRODUCTION
it appears to be essential to produce such a film
if reliable ring performance is to be obtained, and
in fact in all the cases examined, unsatisfactory
ring life was concurrent with absence of a visible
It is hoped that by an examination
transfer film.
of what not to do, a greater understanding can be
obtained of how reliability can be improved still
further.
RESULTS
Figure 1 shows a collection of data from 20 problem
These cases refer to a selection of
cases.
problems encountered by the author over the last
Causes, effects and solutions to the
decade.
problems were obtained by the methodology
illustrated below.
·~
\
In process gas and petrochemical industries
reliable oil-free compres~ors are vital, because
The
of the extremely high downtime costs.
majority of the oil-free compressors used in
these industries are of the horizontal reciprocating type and are generally found to be very
reliable, the pistons being fitted with compression and bearer rings made from PTFE filled with
various inorganic fillers such as carbon, glass
fibre, and molybdenum disulphide, or combinations
In addition there exist composite
of these.
materials consisting of PTFE, carbon, molybdenum
disulphide, compounded with epoxy binders, as
well as alternative filled plastics such as
The
polyimides and polyphenylene sulphides.
compression rings are normally self actuating
and act as sealing elements between the piston
The bearer rings support the
and the cylinder.
weight of the piston and act so as to prevent the
piston contacting the cylinder wall.
The effects and solutions were obtained directly
The causes v1ere
from actual field observation.
derived by relating solution of the problem and
This is best illustrated by
suspected causes.
considering one of the cases in greater depth.
It is found that the rings initially will wear by
up to 0.010" (0.25 mm) and in so doing will
After the
transfer a film to the cylinder wall.
transfer film has been produced the wear rate of
the rings is correspondingly reduced to lower
The mechanism of this film transfer is
levels.
not completely understood and there are various
mechanisms pr~ounded by Arklesf ref. 1,
However,
Richardson re • 2 , and Evans re • 3.
These machines are single stage horizontal, boosting the pressure of bone dry hydroGen gas from 200
p.s.i. (14.1 kg.cm2) to 600 p.s.i. (42.2 kg.cm2).
Monolithic butt-jointed bearer rings and monolithic
self actuating butt-jointed compression rings were
used, fabricated in a carbo-graphite filled PTFE
Initially a variable bearer ring life
material.
of between 200 and 2,000 hours was achieved; also
58
there was a loss of gas throu ghput withi n
a short
perio d of fittin g new rings .
Exam inatio n of the
worn rings revea led that the beare r rings
had
extru ded (i.e. parts of the ring had flowe
d under
the influ ence of press ure and temp eratu re);
also
the comp ressio n rings were found to have
relax ed
so that they were no longe r in conta ct with
the
cylin der wall.
Gas deliv ery temp eratu re recor ds
indic ated that from time to time temp eratu
res of up
to 392°F (200°C) vlere being encou ntere d durin
g
opera tion of the comp resso rs.
The solut ions to the probl ems were
(l)
(2)
(3)
By repla ceme nt of the ring mate rial with
an
epoxy bonded comp osite mate rial bette r able
to cape with a bane dry gas and with highe
r
temp eratu res (ref. 4 and 5).
solut ions.
as:(1)
(2)
By a simil ar metho dolog y (exce pt for cases
4 and
19) cause s, effec ts and solut ions were obtai
ned
for all twent y cases . (Figu re 2).
It shoul d be
noted that for cases 4 and 19, solut ion t-1as
by
other means, i.e. the comp resso rs were conve
rted
to lubri cated dutie s to overcome the probl
ems.
Figur e 2 shows that cause s, effec ts and solut
ions
were able to be categ orise d as follo ws.
(A few
illus trati ve cause s are given in figur e 3.)
By thoro ughly clean ing the water passa ges
of
the cylin der cooli ng and imple menti ng a
regul ar maint enanc e progr am.
This was done
as an attem pt to overcome a high temp eratu
re
cond ition consi dered to be cause d by a
comb inatio n of inade quate maint enanc e and
inade quate desig n of the cylin der cooli ng
syste m.
A.
B.
Modi ficati on was made to the comp ressio n
rings
using a twin ring desig n in which a self
actua ting butt- joint ed inner ring was used
with its gap at 180 degre es to the gap of
a
self actua ting outer ring.
This desig n was
used so as to minim ise leaka ge of the law
mole cular weigh t hydro gen gas past the
comp ressio n rings and there by reduc e any
loss
of gas throu ghpu t.
Effec ts:
Effec ts
(;:;)
Upstr eam system (i.e. the system
prior to the comp resso r).
Use of the comp resso r.
C.
Selec tion of ring mate rials and/o r
desig n of rings .
1.
High ring wear.
High temp eratu re g~v~ng secon dary
effec ts such as ring extru sion.
Ring break age.
2.
3.
4. Loss of comp resso r effic iency .
Solut ions: a. Exclu de extra neous matte r.
b. Remove ovali ty of comp resso r
cylin ders.
c. Use comp resso r corre ctly.
d. Alter ring desig n.
e. Alter ring mate rial.
f. No actio n.
In summary, effec ts and solut ions are
(2)
Incor rect use of the comp resso r.
Incor rect selec tion of ring mate rial and/o
r
desig n.
Discu ssion of Resu lts
As a resul t of the above solut ions, beare
r ring
life was cons isten tly incre ased to a figur
e in
exces s of 15,00 0 hours , exces sive gas temp
eratu
1-1ere nat recor ded, and there was no meas urabl res
e loss
of gas throu ghpu t.
(1)
In this case, cause s were deriv ed
Solut ion by other means
~Lubricate machi ne with an oil
High ring wear.
High temp eratu re, givin g secon dary effec ts
such as ring extru sion.
Lowe ring of gas throu ghpu t.
comp atible with the proce ss.
The range of cause s of the probl ems is shown
in
figur e 3.
It is inter estin g to nate that even
thoug h high ring Hear occur s in all cases
excep t
one, in only appro x. 50% of cases does chang
ing
the ring mate rial solve the probl em.
Solut ions
(1)
Use of comp resso r corre ctly (i.e. imple mentation of a regul ar maint enanc e progr am for
clean ing cylin der water passa ges).
(2) Alter the comp ressio n ring desig n sa
as to
impro ve seali ng.
())_ Alter the ring mate rial so as to comba
t the
bane dry natur e of the gas and the high
temp eratu res encou ntere d.
The resul ts obtai ned are not analy sed to
too great
a depth becau se the natur e of the selec tion
of
cases is nat random and there fore not nece
ssari ly
repre senta tive of ALL probl ems.
Hmvever, it is
felt that certa in c;ncl usion s can be made
as to
good curre nt pract ice so as to achie ve relia
ble
comp resso r opera tion.
Conc lusion s as to good curre nt pract ice.
Cause s were deriv ed by relat ing the solut
ions •1i th
the reaso ns behin d the chang es made to obtai
n the
(1)
59
Grea ter consi derat ion be made as to prov~
s~on
of bette r filtra tion of gas prior to entry
(2)
into a non-lubricated compressor.
The user of a new compressor record all weardown data until such time that a predictable
This not only helps
ring life is achieved.
the user to optimise ring performance but also
helps the 'troubleshooter' in case a problem
develops.
(3)
Care be taken by the user of the machine that
the compressor manufacturer's operating
instructions are observed.
(4)
Careful selection of ring material and/or ring
design be made by the compressor manufacturer.
REFERENCES
(l)
(2)
(3)
(4)
(5)
Arkles B, Theberge J. and Shireson M.
"Wear behaviour of Thermoplastic Polymerfilled PTFE Composites."
Lubrication Engineer Vol. 33, 1, 33-38.
Richardson M.O.w. and Pascoe M.w.
"The Possibility of Reaction between Clean
Iron and Perfluorinated Alkanes."
Wear 18, 426-427 (1971).
Evans D.C.
"The Influence of Abrasive Fillers on the
Wear Properties of PTFE-Based Composites".
A.S.L.E. Conference, Denver (1978).
Tremain R.F.M.
"Overcoming Some Causes of Short Lives of
Rings fitted to Oil-Free Compressors."
A.S.L.E. Conference, Denver (1978).
Maer P.s., Mitchell P.J., Atkins B.R.
"Multiphase Filled-Plastics Piston Rings for
Non-Lubricated Compressors.
Tribology (1973).
6d
Case
Gas
Final Discharge
p,...PR'<lll'P
··--~--~
1
2
-~
/
4
5
6
7
8
9
10
ll
12
13
14
15
16
17
18
19
20
Arr,on
Oxygen
OJ<.;,-gen.
Propylene*
Nitrogen
Nitrogen
Air
Hydrogen
Hydrogen*
Hydrogen*
Propylene
Ethylene
Helium
Hethane
Hydrogen
Air
Carbon
Dioxide
Nitrogen
Carbon
Monoxide
Methane
(psi)
(kg/em"")
Final Discharge
TPmne ·.<~ture
(uF)
(oC)
150
630
600
279
2250
440
383
600
400
375
233
1500
252
150
215
150
130
(10.5)
(44.3)
(42.2)
(19.6)
(158.1)
(30.9)
(26.9)
(42.2)
(28.1)
(26.4)
(16.4)
(105.4)
(17-7)
(10.5)
(15.1)
(10.5)
( 9.1)
255
248
351
189
289
268
338
392
165
320
154
208
268
284
354
330
181
(124)
(120)
(177)
(87)
(143)
(131)
(170)
(200)
(74)
(160)
(68)
(98)
(131)
(140)
(179)
(166)
(83)
150
395
(10.5)
(27 .8)
230
313
141
( 9.9)
252
No. of.
Stages
Configuration
Mean Piston
Sneen
(Ft/min)
2
4
(M/sec
4.
2
2
2
3
Horizontal
Horizontal
Vertical
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Horizontal
Vertical
570
647
388
728
735
722
667
287
500
500
722
644
700
720
613
720
390
(3.6)
(3.7)
(3.1)
(3.7)
(2.0)
(l10)
(156)
2
2
Vee
Horizontal
620
792
(}.2)
(4.0)
(122)
3
Horizontal
Boo
(4.1)
• Gas stated is major constituent.
FIGURE l
61
l
2
5
2
2
l
l
l
2
l
(2.9)
<3-3)
C2.0)
(3.7)
(3.?)
(3.?)
(3.4)
(1.5)
(2.5)
(2.5)
(3. 7)
<3-3)
,
--- ..
/
Case
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Initial
Ring Life
(Hrs.)
Estimated and/or
Ultimate Ring Life
(Hrs.)
2000
350
2000/20000
200
170
500*
500
200
1000
1500
350*
500
1200*
2500
3000
2000
500
100
200
1000
45000
15000+
20000
3000
75000
9000+
8000+
15000
6000+
6000+
8000
8000
3000+
15000
8000+
6000+
8000
8000+
2000
8000+
*
c
c
e
e
d,e
1
1
c
3
A
1
1
1
1
1,2,4
1
1,2
1
1,2,4
1,4
1
1,2
1
1
1
1
1
c
A
A
B,C
A
B
A
B
A,C
c
B
A,C
c
c
A,B,C
c
Solution by
other means
oL_
e
a
a
c,d,e
a,b
f
a
c
a,d
e
f
a,d,e
e
e
e
cL
Applies to first stage only.
~
A. Ups"tr•eam system.
B.
Inferred
Cause
Solution
Effect
Incorrect use of
compressor.
c. Incorrect selection of
ring material and/or
ring design.
Solution
Effect
1.
High ring wear.
2.
High temperature .
3· Ring breakage.
4. Lack of sealing
efficiency.
FIGURE 2
62
a.
Exclude extraneous
matter.
b.
Remove cylinder
ovality.
c.
d.
Use compressor
correctly.
Alter ring design.
e.
Alter ring material.
f.
No action.
Solution by
other means
..!...
Lubrication .
A Few Illustrative Causes
;...
Upstream System
B.
Use of Compressor
·---·
Ingression into compressor of
rust, concrete, and welding
debris.
Partial polymerisation of gas.
Gas temperature too high.
Lack of maintenance of
cylinder cooling.
Dynamiting in the vicinity
of the compressor.
c.
Selection of Ring Material
and/or Design
Incorrect ring design.
Incorrect ring material
selection in relation to the
dryness of the gas (Ref. 4, 5)
and incorrect material
selection in relation to the
gas temperature.
Incorrect material selection
in relation to the nature of
the gas.
FIGUHE 3
'·
63