UNIVERSITY OF NIŠ
The scientific journal
FACTA UNIVERSITATIS
Series: Mechanical Engineering Vol.1, No 6, 1999 pp. 711 - 719
Editor of series: Nenad Radojković, e-mail: radojkovic@ni.ac.yu
Address: Univerzitetski trg 2, 18000 Niš, YU
Tel: +381 18 547-095, Fax: +381 18 547-950
http://ni.ac.yu/Facta
ENDURANCE TEST OF MICRO PUMPS FOR BEARINGS
LUBRICATION IN THE TURBOJET ENGINES
UDC 621.176, 62-728
Dragoljub Vujić
Military Institute, Katanićeva 15, 11000 Beograd, Yugoslavia
Abstract. This paper presents the procedure for an endurance test conducted on the
micro pumps for bearings lubrication in turbojet engines. The endurance test was
carried out on the test rig installation using a procedure for a working regime
simulation of the micro pumps. Analysis and evaluation of the investigation results are
performed on the basis of the mathematical statistics methods.
Key words:
Turbojet engine, lubrication system, micro pumps, endurance test.
1. INTRODUCTION
Endurance investigations are usually conducted in the development phase of new
equipment, or when materials or production technology are changed. The preconditions
and load spectra used in the investigation are defined by the designer. He prescribes a
duration time of investigations, which are usually conducted up to the failure of the
component (damage appearance), or accomplished through a number of equivalent cycles
defined in the program investigations.
This paper presents a procedure and part of the results of an endurance test of the
micro pumps for bearings lubrication in turbojet engines. Investigations were permfored
under laboratory conditions using a test device for simulation working regime of the
micro pumps. The test lasted 1000 hours.
2. LUBRICATION SYSTEM
The lubrication system is self-contained with the engine and comprises a pressure and
scaving system. The pressure system comprises two distinct circuits: the main pressure
circuit which is supplied by the main oil pump to serve the front bearing, bevel gear box
Received December 28, 1998
D. VUJIĆ
712
and the accessory drive gearbox. In addition to this a separate (metered) circuit is
supplied from the main system and delivered as a measured quantity from each of two
micro pumps, one to supply the rear bearing and one the centre bearing (Fig. 1). This oil
is finally burned in the jet pipe.
centre main
bearing
frornt main
bearing
oil
tank
oil
pump
oil pressure
transmitter
micro
pump
micro
pump
Fig. 1. Schematic representation of the oil lubrication system
((((
SCAVENGE PUMP
TO REAR MAIN
BEARING
TO CENTRE MAIN BEARING
RETURN TO
TANK
INLET FROM
TANK
PRESSURE
PUMP
FILTER
INLET TO SCAVENGE
PUMP
TO ACCESSORY GEARBOX
TO FRONT MAIN
BEARING, BEVEL BOX
AND OIL PRESSURE
TRANSMITTER
MICRO PUMP
MICRO PUMP
SUCTION FROM
TANK
OIL PRESSURE
PUMP
METERED PRESSURE
SCAVENGE AND
DRAIN
PRESSURE
Fig. 2. Schematic view of the oil pump.
The oil pump is located on the accessory gearbox at the base of the air intake casing.
The main oil pump unit (Fig. 2) comprises a separate gear type pressure and scavenge
pump. A poppet type static sealing valve fitted adjacent to the pressure pump outlet,
Endurance Test of Micro Pumps for Bearings Lubrication in the Turbojet Engines
713
closes the pump delivery port when the engine is stationary to prevent oil seepage into the
engine. A spring loaded piston type relief bypasses oil back to the pressure pump inlet to
maintain the supply pressure at the required level.
Micro pumps are mounted on the oil pump cover. Essentially each pump (Fig. 3)
consists of a positively driven pumping piston and spring loaded metering piston.
Movement is transfered to the pumping piston by a yoke that is actuated by a cam. The
metering piston located opposite each pumping piston is spring loaded and housed within
a common ported cylinder and is operated by the force transmitted through the oil to
uncover the delivery port.
Fig. 3. Cross-section of micro pump
3. INVESTIGATION PROGRAM
The investigation program was performed on the basis of aircraft "flight profile" and
contained working regime simulation of the micro pumps for three flight sortie patterns:
low-low, high-low and training mission. Investigation is conducted in two phases on two
arbitrarily selected micro pumps. Each phase lasted 500 hours.
During the test the value of certain parameters was adjusted: speed of the oil pump
driving shaft, inlet and outlet micro pumps pressure and regime duration time for each of
the mentioned missions. Total duration time testing for all three missions is presented as
one cycle. Flow checking is permfored after each of 10 cycles, after first phase of 500
hours, before starting the second phase and at the end of the test (after second phase).
Special attention is paid to prevent oil leakage on the outside of the micro pumps
surface during testing. Outlet pressure of 3 �� 0.5 bar of the micro pumps was achieved by
closed valve (item 16 and 17 on the Fig. 4) at 2.07 bar inlet pressure and 65 �� 50C
working fluid temperature.
D. VUJIĆ
714
t
EM
n
4
7
9
11
13
10
12
14
19
17
16
7
17
19
15
16
5
20
15
7
MPl
MPd
1
2
3
UF
PP
CP
NV
PF
PF
8
PV
6
18
5
18
Fig. 4 - Shematic view of test device
1 - electromotor, ZIM 250 MKK , !S, type, r.p.m = 50 to 3260 min-1, 2 - electromotor r.p.m gauge,
3 - digital r.p.m indicator, 4 - oil pump (PP-pressure pump, CP-scavenge pump, PV-relief valve,
NV-non-return valve, UF-suction fir, PF-outlet pressure filter), 5 - gauge, V = 10 cm3, 6 - oil
supply tank, 7 - pressure gauge, 0 to 10 bar, class 1,160, 8 - heater, 2��1 kW, 9 - contact
thermometer, 10 - level indicator, 11 - electric level, 12 - filler hole, 13 - draining valve, 14 - ball
valve, 15 - micro pump (MPl-micro pump left, MPd-micro pump right), 16 - ball valve,
17 - adjustable valve for small flows, 18 - safety valve 744-0300, 19 - oil tank, V = 2 dm3,
20 - multiplication gearbox, MR 2.010, i = 2, type
Dimensional checking of selected parts and subassemblies of the micro pumps after
accomplished a fixed number of cycles was defined in the test procedure.
4. ENDURANCE TEST
Before beginning the investigations, a dimensional inspection of particular parts was
permfored and the functional characteristics of the micro pumps was checked. The
dimensional inspection consisted of disassembling and measuring characteristic
dimensions of particular items and subassemblies of the micro pumps. Fig. 2 presents
only two measured places (points M.P.1 and M.P.2) which are most interesting for
analysis in this investigations. After accomplishing satisfactory values according to
regulation on desired micro pumps quality, the endurance test started. In table 1 are given
prescribed values of clearance at the points M.P.1 and M.P.2 and measured values during
investigation. The functional relationship between a flow and realised clearance will not
be considered in detail here.
A schematic view of the test device specially designed for this investigation, is
presented in Fig. 4. For investigation synthetic engine oil MOBIL JET OIL II was used
according to MIL-L-23699B. Oil temperature during investigation is controlled between
650C to 700C.
Endurance Test of Micro Pumps for Bearings Lubrication in the Turbojet Engines
715
Table 1. Clearance values at the measure points M.P.1 and M.P.2
Measured values (All dimensions in [mm])
Before beginning
of investigations
After first phase
of 500 hours
After second phase
of 500 hours
Measured
points
clearance
Prescribed
values
Pump No1 Pump No2 Pump No1 Pump No2 Pump No1 Pump No2
M.P.1
0,002-
0,005
0,004
0,004
0,004
0,004
0,005
0,005
M.P.2
0,002-
0,005
0,004
0,004
0,004
0,004
0,004
0,005
The simplest measured parameter, and at the same time, the most reliable indicator of
work and behaviour of the pumps during investigation was a flow. By measuring the flow,
and dimensionally checking particular parts of the micro pumps at a fixed number cycles,
the investigator was able to controll all the investigation process.
For practical reasons, here we shall present polygon flow distribution (Fig. 5) and
parameters which defined its distribution.
97,5
101,5
105,5
109,5
113,5
0
2
4
6
8
F
re
q
ue
nc
y
[-]
Mean of the class [cm3/h]
Pump No.1
r.p.m = 1860 min-1
97,5
103,5
109,5
115,5
121,5
0
2
4
6
8
F
re
q
ue
nc
y
[-]
Mean of the class [cm3/h]
Pump No.2
r.p.m = 1860 min-1
290,5
296,5
302,5
308,5
314,5
0
2
4
6
8
10
12
F
re
q
ua
nc
y
[-]
Mean of the class [cm3/h]
Pump No.2
r.p.m = 5000 min-1
294
299
304
309
314
0
2
4
6
8
10
F
re
q
ua
nc
y
[-]
Mean of the class [cm3/h]
Pump No.1
r.p.m = 5000 min-1
Fig. 5. Polygon of empirical flow distribution
D. VUJIĆ
716
5. MATHEMATICAL CALCULATIONS USING STATISTICAL METHOD
By using mathematical statistics method the main parameters which defined flow
distribution are calculated. By well known procedure for statistical data processing
measured values of the flow are classified into classes. The table 2 presents values of the
delivery flow and frequencies for two characteristic speeds
n = 1860 rpm and
n = 5000 rpm.
Table 2.Delivery flow and frequencies
Pump No.1
Pump No.2
Delivery flow
[cm3 /h]
Mean of the class
(arithmetic mean)
xi[cm3/h]
Frequency
[-]
Delivery flow
[cm3 /h]
Mean of the class
(arithmetic mean
xi [cm3/h]
Frequency
[-]
Driving speed
n = 1860 rpm
96 - 99
97,5
1
95 - 100
97,5
8
100 - 103
101,5
5
101 - 106
103,5
7
104 - 107
105,5
8
107 - 112
109,5
7
108 - 111
109,5
7
113 -118
115,5
2
112 - 115
113,5
4
119 - 124
121,5
1
Driving speed
n = 5000 rpm
292 - 296
294
1
288 - 293
290,5
1
297 - 301
299
6
294 - 299
296,5
2
302 - 306
304
6
300 - 305
302,5
12
307 - 311
309
9
306 - 311
308,5
9
312 - 316
314
3
312 - 317
314,5
1
To process the results statistically, let us denote with the
x1,
x2, ∙∙∙,
xn, flow values
noted during investigation and with
f1,
f2, ∙∙∙,
fn frequencies of its appearance.
(
f1 +
f2 + ∙∙∙ +
fn =
N is measuring number).
From parameters representing centre distribution, we quote here the arithmetic mean,
empirical dispersion and standard deviation.
The arithmetic mean
x and empirical dispersion
s2can be, respectively, presented in
following form:
��
��
=
=
-
=
=
n
i
i
i
n
i
ii
xxf
N
s
fx
N
x
1
2
2
1
)
(
1
1
Positive square root from dispersion, known as standard deviation �� (for great values
of N, according to law of great numbers,
S is a little different from �� ).
For a detailed description of flow frequency distribution are introduced according to
[2,3,4] moment coefficient of skewness
A
K and moment coefficient of kurtosis
E
K in
form:
3
,
4
4
3
3
-
��
=
��
=
s
K
s
K
E
A
where
)4,3
( ,)
(
1
1
=
-
=
��
��
=
r
xxf
N
r
i
n
i
i
r
Endurance Test of Micro Pumps for Bearings Lubrication in the Turbojet Engines
717
are central moments of higher order (the third and the fourth order).
For comparison of dispersions of the different statistical assemblies, a coefficient of
variation
V
K is introduced, as relative dispersion characteristics in form:
%
100
x
s
x
s
KV
=
=
Calculated parameters of dispersion centre for both investigation phases are presented
in table 3.
Table 3. Parameters of centre dispersion
Parameters of centre dispersion
pump No.1
n = 1860 rpm,
n = 5000 rpm
pump No.2
n =1860 rpm,
n = 5000 rpm
Arithmetic mean
x
106,78
305,40 104,94
304,58
Empirical dispersion
s2
18,84
29,04
44,00
17,43
Standard deviation ��
4,34
5,39
6,63
4,17
Moment coefficient of skewness
A
K
1,42
2,00
1,49
1,52
Moment coefficient of kurtosis
E
K
- 0,76 - 0,87 - 0,36 - 0,24
Coefficient of variation
V
K
4,06
1,76
6,32
1,37
In order to better understand the investigation results, a graphical illustration of
coefficients
A
K and
E
K is given in the Fig. 6.
Moment coefficient of skewness
Moment coefficient of kurtosis
Fig. 6. Graphical illustration of coefficients
KA and
KB .
At symmetrical distribution
0
=
A
K
. If the moment coefficient of skewness is negative
(
0
<
A
K
), the frequency curve is skewed to the left. If moment coefficient of skewness is
positive (
0
>
A
K
), the frequency curve is skewed to the right. It is obvious, that
distribution is more asymmetrical, if the absolute value of the moment coefficient of
skewness is greater.
If
1,0 |
| 0
<
<
A
K
, no asymmetry
25,0 |
| 1,0
<
<
A
K
, asymmetry is small
5,0 |
| 25,0
<
<
A
K
, asymmetry is medium
|
| 5,0
A
K
<
, asymmetry is strong.
D. VUJIĆ
718
6. ANALYSIS OF INVESTIGATION RESULTS
State of the parts and subassemblies was satisfactory, without visible damage trails.
Only usual small wear is noticeable out moving parts as consequence of long-term pumps
operations. This wear did not change the pumps' functional characteristics. Measured flow
values remained within permissible boundaries during all investigation time. Clearance at
the end of investigations, measured on the checking points M.P.1 and M.P.2 remained
also in prescribed boundaries. At both micro pumps, there was evidently strong
asymmetry of flow curve because coefficient (
5,0 |
|
>
A
K
) (see [2] ).
Asymmetry at the pump No.1 is somewhat greater (
00,2 |
|
>
A
K
) than asymmetry at the
pump No.2 (
52,1 |
|
>
A
K
) with speed of 5000 rpm. At both micro pumps, moment
coefficient of kurtosis is negative. At pump number No.2, absolute value of the moment
coefficient of kurtosis is something smaller than at pump No.1, that the curve is more near
to normal (Gaussian curve) distribution (
0
=
E
K
). On the basis of coefficient of variation
V
K
it can be concluded that if speed is higher (
n = 5000 rpm), at both pumps it is
evidently greater homogeneity of flow values (
76,1
=
V
K
i 1.37) than at smaller speed
(
n = 1860 rpm) (
06,4
=
V
K
and 32,6 ).
7. CONCLUSION
By using mathematical statistics methods, the main parameters which represented flow
distribution during micro pumps testing can be calculated. The distribution type is most
probably Erlang's as a special case of exponential distribution. Coefficients and this
distribution can be permfored on the basis of further pumps investigations using presented
procedure up to failure pumps appearance or on the basis of failures number during life of
the pumps.
REFERENCES
1. Grant L.E., Leavenworth S.R., (1998),
Statistical Quality Control, Mc Graw-Hill Book Company,
Singapore
2. Vukadinović S., (1973),
Elements of probability theory and mathematical statistics, Privredni pregled,
Beograd
(in Serbian)
3. Elishakoff I., (1983),
Probabilistic methods in the theory of structures, John Wiley&Sons, Inc., USA, New
York
4. I-DEAS test (User's guide), (1990), Structural Dynamics Research Corporation, Milford, Ohio, USA
5. Vujić D., (1998),
Analysis of piston geometrical parameters effects on flow in micropumps for lubrication
of turbojet engine bearings, Naučno-tehnički PREGLED, Vol.XLVIII, No.2, pp.22-25. (
in Serbian)
6. Vujić D., (1996),
Endurance test of the micro pumps, Naučno-tehnički PREGLED, Vol. XLVI, No 4-5,
pp.98-102.
Endurance Test of Micro Pumps for Bearings Lubrication in the Turbojet Engines
719
TEST IZDRŽLJIVOSTI MIKRO PUMPI ZA PODMAZIVANJE
LEŽAJEVA TURBOMLAZNIH MOTORA
Dragoljub Vujić
Rad prikazuje proceduru testa izdržljivosti sprovedenu na mikro pumpama za podmazivanje
ležajeva turbomlaznog motora. Test izdžljivosti izvršen je u laboratorijskim uslovima, na ispitnom
uređaju, simuliranjem režima rada mikropumpi. Analiza i ocena rezultata ispitivanja izvršena je
na osnovu metoda matematičke statistike.
Ključne reči:
Turbomlazni motor, sistem za podmazivanje, mikro pumpe, test izdžljivosti.
