-
MTZM O T O RT E C H N I S C H E Z E I T S C H R I F T
Measures to Limit the Latent Operational Danger of Large Marine
Diesel Engines (above 2.25 MW)
Offprint for
- 66432 Blieskastel . Saarland . Germanyel. + 49 (0) 6842 / 508
- 0 Fax + 49 (0) 6842 / 508 - 260
MTZ Motortechnische Zeitschrift 62 (2001) Vol. 7/8 & 12
Friedr. Vieweg & Sohn Verlagsgesellschaft mbH, Wiesbaden
-
1 Introduction
Large marine diesel engines, which in thiscontext means all
those diesel enginesthat are subject to monitoring in accor-dance
with the SOLAS safety regulationsfor fire precautions, can be a
source of
considerable danger during their opera-tion. Therefore,
continuous safety moni-toring of the operating conditions is
re-quired in order to avoid both primary andsecondary damage. The
operational dan-ger of large diesel engines can lead to
thedevelopment of serious fires or to sec-
ondary damage, with the consequencethat the ship is no longer
manoeuvrable.
2 Latent Operational Danger
Two main areas of danger can be deter-mined: breakage of machine
components as a
result of unexpected, sporadic mechan-ical overstress
lack of lubrication on journal bearingsand sliding surfaces.
Part 1 of this article by Schaller Automation examines the
internationalissues involved in establishing and complying with
safety regulationsand the necessary technical measures to ensure
the safe operation oflarge diesel engines.
Measures to Limit the Latent Operational Danger of Large Marine
Diesel Engines (above 2.25 MW) Part 1
DEVELOPMENT
-
3MTZ 12/2001 Jahrgang 62
2.1 Sporadic Mechanical Overstress
A particularly dangerous situation oc-curs when the engine
over-revs as a re-sult of insufficient speed control, for ex-ample
if the main drive coupling breaks.Instant cut-off of the fuel
injection with-in a fraction of a second can prevent ex-cess speed
and therefore limit or avoiddestructive acceleration forces of
themoving masses. State-of-the-art moni-toring systems can be made
sufficientlysensitive and quick to react, and withadded redundancy
their use becomesparticularly reliable.
2.2 Lack of Lubrication
Contrary to the method of resolving prob-lems of mechanical
strength in machineparts, which, with the aid of modern struc-tural
analysis and computation methods,results in a high level of
component relia-bility, determining the tribological
charac-teristics of lubrication oils relies to a greatdegree on
empirical studies. A clear defini-tion of when the lubrication is
starting tofail or when a lack of lubrication is evidentis not yet
generally available, in contrastto the application methods for
determin-ing the strength of materials as mentionedin Section
2.1.
The state-of-the-art monitoring systemsavailable today are
limited to monitoringthe physical after-effects of failed
lubri-cation, as manifested in the rise in temper-ature of a
particular sliding surface as a re-sult of friction. A very
promising methodfor monitoring bearings is the recording
ofthermo-currents generated by the differ-ent metals of the sliding
surfaces (e.g.shaft/bearing). The system developed bySchaller
Automation (BEAROMOS = Bear-ing Overheating Monitoring System)
isnow being perfected for application inpractical test cases.
Deficient lubrication results in frictionthat generates heat,
which manifests itselfin an increase in temperature on the slid-ing
surfaces. Lubricating oil evaporatingfrom an overheated part
re-condensesagain and produces a very dangerous phe-nomenon: the
generation of explosive oilmist. Large engines with
substantialcrankcase spaces are susceptible to this la-tent
operational hazard. The followingwill describe the dangerous
phenomenonof a crankcase explosion as the result oflack of
lubrication and oil mist formation,Figure 1.
3 Oil Mist as a Dangerous Element during the Operation ofLarge
Diesel Engines
3.1 History
3.1.1
In 1947, the occurrence of a dramaticcrankcase explosion in a
ship [1] led Britishinstitutes to conduct scientific
investiga-tions into the phenomenon. The findings,published in the
mid 1950's [2, 3, 4], stillrepresent the most important
scientificstudy of the operational hazard involvingcrankcase
explosion, especially in largediesel engines.
The international safety regulatory body,the International
Maritime Organisation(IMO), with 156 member states, is an
sub-organisation of the UN and has its head-quarters in London. It
was established in1982 by the Inter-Governmental
MaritimeConsultative Organisation (IMCO), whichassumed its duties
in 1948.
Through the IMO, the SOLAS (Safety of Lifeat Sea) Regulations
were launched withreference to fire precautions. Chapter
II-1Construction Part E Additional require-ments for periodically
unattended ma-chinery spaces, Regulation 47 correlatesto the fire
hazard developing fromcrankcase explosions.
3.1.2The intention of these regulations, which
were introduced in the 1960's, was to im-prove the safety of
marine engines and re-quire monitoring by means of oil mist
de-tection as a means of avoiding crankcaseexplosions. The
regulation refers in partic-ular to all internal combustion
engineswith an output of more than 2.25 MW orthose with a cylinder
greater than 300mm. Further protective measures men-tioned in the
relevant SOLAS regulationsfor the prevention of crankcase
explosionsare not clearly defined (for example: en-gine bearing
temperature monitors orequivalent devices), and a
superficial,merely cost-oriented interpretation is nota safe means
of preventing fire caused bycrankcase explosion.
The supervision and implementation ofthe SOLAS regulations is
largely the re-sponsibility of national classification soci-eties.
Of these, the major ones formed anassociation in 1969 called IACS
(Interna-tional Association of Classification Soci-eties). IMO and
IACS (as an intermediary),Figure 2, compile and control the
interna-tional regulations for ship safety in a de-fined
cooperation [5].
3.1.13
In recent years, as a consequence of dra-matic shipping
catastrophes involving sig-
Figure 1:A crankcase explosion can destroy an engine down to
scrap
-
nificant and costly environmental dam-age as well as the loss of
material and unprecedented loss of human life, more effective
safety regulations wereagreed upon on an international level,with
the intention of avoiding furthercatastrophes.
3.1.4
These include two agreements in particu-lar: The Marpol
Agreement (PSC = Port
States Control, Internet: Equasis.org), inwhich the port member
states areobliged to inspect the docking shipsand verify that they
comply with thevalid safety regulations. If deficienciesare
discovered, the ships are to be pre-vented from embarking until
they com-ply with the safety regulations
The ISM Code (International Safety Man-agement Code), which has
the primaryobjective of maintaining safe shippingoperation and the
prevention of marinepollution by ensuring that operators ofships
fulfil the obligations and responsi-bilities determined by the
Code.
The Code establishes safety managementobjectives and requires a
safety manage-ment system (SMS) to be established bythe Company,
which is defined as theship owner or any person, such as themanager
or chartering agent, who has as-sumed responsibility for operating
theship. The Company is then required to es-tablish and implement a
policy forachieving these objectives. This includesproviding the
necessary resources andshore-based support. Every company
isexpected to designate a person or per-sons ashore having direct
access to thehighest level of management. The proce-dures required
by the Code must be docu-mented and compiled in a Safety
Man-agement Manual, a copy of which is to bekept on board.
4 Technical Measures for Ensufing Safe Ship Operation
Apart from measures that guarantee thestability of the ship's
hull, all equipmentthat is necessary for the safe operation ofthe
ship must be fully functional.
4.1
The ship can only manoeuvre if the ship'spropulsion system and
the rudder systemare in working order. For the majority of
modern ships built in the last few decades,diesel engines were
installed to producemechanical power for propulsion, as wellas to
provide electrical power to drive aux-iliary equipment. Auxiliary
equipment in-cludes such things as the hydraulics for therudder
activation or in some cases thediesel-electric propulsion of the
ship.A very important measure for the safe op-eration of the ship
is the preservation ofthe diesel engines operation. For this
rea-son, the diesel engine must not be al-lowed to suffer severe
damage and anom-alies should be reduced to a minimum bymeans of
early recognition. This would al-low a simple repair to be carried
out onboard or would facilitate assistance fromland to sea.
4.2
The classification societies are currentlyestablishing rules for
redundant diesel en-gine power units, especially to secure
thefunction of the ship's propulsion systemand the function of the
rudder.
5 Oil Mist as a Damage Indicator
Oil mist is a dangerous element but, ifrecognised early, it can
also serve as asuitable indicator that measures are re-quired to
prevent the marine diesel en-gine from suffering damage.
Today's oil mist detectors, Figure 3, Figure4 and Figure 5,
utilize a light path for therecognition of the passing oil mist.
The oilmist itself causes clouding in the lightpath [6]. Highly
concentrated and explo-sive oil mist (air/oil droplet mixture)
pro-duces a very strong cloudiness (light ab-sorption). For the
early detection of oilmist, such as that generated by
frictionaldamage as a result of failed lubrication,the light path
measuring system has to re-act with high sensitivity. This results
in adiscrepancy due to the fact that, duringnormal engine
operation, oil mist of a less-er concentration also develops in
areas inwhich the evaporation temperature of theoil is reached
(above 230C). This conditionleads to false alarms, which trigger
un-necessary automatic engine shut-downs.However, this depends on
how the oil mistdetection has been integrated into the en-gines
safety system.
It is obvious that only an immediate en-gine stop or at least a
reduction in power(slow-down) to reduce the ongoing fric-tion will
limit the damage efficiently. Atpresent, no solution to this
problem can beproposed, since there is still a lack of fun-damental
studies.
5.1 System Configuration for Efficient Safety Equipment
An additional question with regard to theapplication of oil mist
detectors for engine
DEVELOPMENT Large Marine Diesel Engines
4 MTZ 12/2001 Jahrgang 62
Figure 2: Intermediarymodel of the classifica-tion societies
[5]
-
5MTZ 12/2001 Jahrgang 62
protection is that of the right system con-figuration for
efficient engine safetyequipment based on oil mist detection(OMD).
As yet, the classification societies(Section 4) have not
established the ruleswithin the framework of IACS,.
Due to the pressure of cost reduction in thesupply industry for
ships and marineequipment, the lack of unequivocal regu-lations
leads to inadequate solutions forsafety related to oil mist
detection. It isclear that, during normal trouble-free op-eration,
safety equipment is irrelevant andonly becomes necessary at the
momentwhen the emergency occurs.
In addition, a number of other non-lubri-cation-related
anomalies in the diesel en-gine can be recognized, for example
whenthe cooling of pistons is failing and thetemperature increase
causes an exception-al oil mist generation that is ultimately
de-tected.
6 OMDEA, Oil Mist Detection Efficiency Approval
Following a meeting attended by repre-sentatives from
classification societies,OMD manufacturers and scientists,Schaller
Automation established the pro-ject OMDEA with the aim of
developingimproved fundamentals for controlled en-gine safety.
Within the framework ofstringent safety regulations for ships,
en-gines are also required to have certifica-tion known as the Type
Approval Certifi-cation (TAC). Included in the TAC are testsfor
safety equipment, such as measuresagainst crankcase explosions.
6.1 OMDEA Certificate
Schaller Automation is currently develop-ing the fundamentals
for making the cor-responding certification applications tothe
classification societies. For an OMDEAcertificate to be issued, the
scientificallybased principles of oil mist generation re-sulting
from lack of lubrication must bepresented.
Schaller Automation commissioned theInstitut fr
Maschinenkonstruktionslehreund Kraftfahrzeugbau (mkl) at the
Univer-sity of Karlsruhe, Germany, under the di-rection of Prof. A.
Albers, to scientificallyformulate these basic principles. The
find-ings from the experiments carried outwere so encouraging that
a second seriesof tests has been added to the program.
Figure 3: Oil mist detector installed on a crosshead two-stroke
engine. Power output 9 MW (output up to 70 MW per unit can be
realized)
Figure 4: Oil mist detector installed on a four-stroke diesel
engine, 2,4 MW, up to 50 MW per unit can berealized
-
The findings of the research will be pre-sented in Part 2 of
this article. The resultsof test series conducted by the
companyFMC, Fiedler Motoren Consulting KielGmbH, Germany, will also
be published inPart 2.
7 Concluding Remarks
In an internet search to find informationon marine accidents
caused by diesel en-gines, the term Crankcase Explosion ap-peared
more than one thousand times. Atpresent, the information is being
scruti-nized within the OMDEA project.
Schaller Automation has sponsored a neu-tral Internet Forum,
www.dieselsecurity.org, with the aim of raising the awarenessof
diesel engine safety problems on a glob-al scale among those
involved. Further-more, it aims to encourage the exchangeof
experience relevant to the improve-ment of safety measures. This
also appliesby analogy to land-based diesel powerplants.
In Part 2 (planned in MTZ 12/2001), a reportwill present
scientifically attained resultsof tests carried out to enable a
better as-sessment of the physical phenomena ofsliding surfaces and
journal bearings run-ning under poor lubrication conditions.
DEVELOPMENT Large Marine Diesel Engines
6 MTZ 12/2001 Jahrgang 62
Figure 5: Classic oil mist detector provided with siphon block
suction system for each crankcase compartment
-
7MTZ 12/2001 Jahrgang 62
Part 1 [7] of this article by Schaller Automation pointed out
that there are interna-tionally controlled safety regulations on
this subject. However, these regulationsare not specific and their
implementation is difficult, since there is a lack of pre-cise
technical and verifiable definitions. Part 2 describes the efforts
of SchallerAutomation to make the fundamental safety measures
achievable in a practicablemanner and to allow the efficiency of
these measures to be confirmed withoutthe need for such
compromising concessions as where practicable [8]. In orderto
create the basis for achieving this aim, tests and extensive
measurementswere performed at the Institute for Machine Design and
Vehicle Construction ofthe University of Karlsruhe (TH). The
findings of this research show that a definiteimprovement in engine
protection measures is possible, and positive results canbe
achieved by the controlled application of these measures.
DEVELOPMENT
Measures to Limit the Latent Operational Danger of Large
MarineDiesel Engines(above 2.25 MW) Part 2
-
DEVELOPMENT Large Marine Diesel Engines
8 MTZ 12/2001 Jahrgang 62
9 Introduction
Part 2 of this article described how SchallerAutomation is
pursuing efforts to obtainmore tangible information on the extent
ofdamage to large diesel engines on a world-wide scale.
In spite of the support provided by Clas-sification Societies,
these efforts haveunfortunately not been very successful.Therefore,
consideration of this subjectdepends more or less on calculated
specu-lation supported by personal practicalexperience and
empirical evaluations.From information and stored data current-ly
available, it was possible to derive that,related to the service
life of all engines, theoccurrence of crankcase explosions lies
inthe lower single percentage range.
Without the mandatory monitoring reg-ulation for large marine
diesel engines, thepercentage of crankcase explosions
wouldcertainly be higher. However, a significantnumber of these
safety-enhancing installa-tions must be discredited, Figure 7
[9].
This example of an installation forwhich the client did not
utilize the urgent-ly needed design support reveals deficien-cies
that make the correct functioning ofthe OMD system (Oil Mist
Detection)impossible. The sampling pipes were fittedhorizontally
and parallel to the engine andwill therefore fill up with the
fallout oil ofthe oil mist. What is even more disturbingis the
sagging of the pipe that is formed bythe hose connection (Figure
7), which willcertainly fill up with oil and impede theextraction
at that point. In addition, thelikelihood of condensed water
beingtrapped within the sagging pipe will makethe safety-relevant
function unworkableand cause long-term damage to the system(rust
deposits).
The example shows that only a com-plete OMD system adapted for a
specifictype of engine should be designed, deliv-ered and
commissioned by one and thesame organization, the manufacturer.
Inorder to improve the OMD protection mea-sures for large diesel
engines, SchallerAutomation initiated the OMDEA (Oil MistDetection
Efficiency Approval) Project. PartI, Sections 6 and 6.1 [7]
contains basic testsrelevant to oil mist formation.
10 Proposal for the Structureof OMDEA Certification
andRequirements for the DerivedExperiments
10.1 Efficiency of an OMD SystemThree efficiency categories are
recom-mended:
10.1.1Protection Category 1The minimum efficiency required for
aprotection system encompasses only theprevention of explosions. It
fulfils therequirements of SOLAS Regulations forfire precautions
(Part 1 [7], Section 3.1,Paragraphs 3 and 4)
10.1.2 Protection Category 2Evaluation of oil mist development
in thecrankcase by the OMD system to preventsevere frictional
damage by means of dif-ferentiated automatic intervention in
theengine operation. It by far surpasses theminimum SOLAS
requirements.
10.1.3 Protection Category 3 Damage localization based on
Protection Cat-egory 2. This enables the quick location of
theaffected compartment and recognition of thedamage. Damage that
has not been locatedsuccessfully can result in the affected
engine
being restarted by the personnel, thusseverely aggravating the
condition.
10.2 Assessment Criteria forthe Effectiveness of an
OMDsystemThree criteria are paramount for assessingthe
effectiveness of an OMD system:
10.2.1 Criterion 1: Sensors and Evaluation In the case of
developing damage, suffi-cient sensor sensitivity must be
availablein relation to the generated oil mist con-centration
(opacity, %OP), and a reliablesoftware evaluation for the indicator
func-tion of the oil mist must be ensured.
10.2.2 Criterion 2: Immunity to false alarms An oil mist warning
signal should not betriggered if no frictional damage is devel-
9 Introduction
Figure 7: Incorrect installation of an oil mist detector
VISATRON VN215with individual suction pipe connected to each
compartment for localizati-on of damage1. VISATRON VN 2152. Valve
box with reed valves for selection of suction pipes according to
asearch-run algorithm. Indication of damage compartment (red dots
in displaywindow)3. Individual suction pipes4. Hose connection
between the suction pipes and pipe elbows inserted in valvebox 25.
Sagging hose in the sampling system6. Connection to the matching
compartment
-
9MTZ 12/2001 Jahrgang 62
oping. This criterion is even more impor-tant for the acceptance
of an OMD systemthan Criterion 1.
10.2.3 Criterion 3: Availability of the protectionsystem The
reliability of the OMD system must beproven, taking into
consideration marineenvironment conditions, the direct
instal-lation on the engine and the obligatory
tests (vibration, temperature, resistance tohumidity, high
EMC).
10.2.4 The Levels of OperationalSafety System Safety Level 1:
The system con-trols itself and a special alarm signal isactivated
when the protection function forascertaining oil mist is
malfunctioning(low-cost version) System Safety Level 2: In addition
to
System Safety Level 1, a second systemprovides redundancy. For
both safety lev-els, a dedicated worldwide maintenanceservice must
guarantee the availability ofthe systems.
10.2.5 Influence of an OMD System on Engine OperationAt present,
there is no clear decision onwhether an OMD should react by
activat-ing STOP or SLOW DOWN of the engine inthe event of a lack
of lubrication. A solu-tion to this point must be found,
especial-ly in the case of nautical emergency situa-tions. The
ship's master should be support-ed by sensitive software to provide
an aidto decision-making.
11 Experimental Investigations
11.1 Tribological Tests onLubricated Sliding Surfacesand Oil
Mist Simulation Tests inLarge EnginesExtensive tests were conducted
in order tofind out more about the oil mist resultingfrom
frictional damage caused by insuffi-cient lubrication. For the
OMDEA project,the aim is to study which methods arerequired for the
purpose of simulating oilmist development and oil mist behaviourin
the crankcase.
In order to obtain meaningful knowl-edge about oil mist
formation, two differ-ent experimental rigs for sliding
bearingswere utilized: A specially designed test rig for a
radialsliding bearing A linear sliding guide for the simulationof
the piston/cylinder function, in theform of a rotating disc with a
lubricatedguide block
11.2 Examination of RadialBearings11.2.1 Test Bench and
Measurement MethodThe experiments were conducted on thesmall,
Figure 8, and the large, Figure 9,radial bearing test bench of the
Institutefor Machine Design and Vehicle Con-struction of the
University of Karlsruhe(TH) by Prof. Dr. A. Albers. The set-up
ofthe small bearing test bench is shown inFigure 10. The test
bearing on which theloading force is applied by means of ahydraulic
cylinder is located between thetwo support bearings, which hold
theshaft radially.
The small bearing test bench enablesexperiments with 180 half
plain bearingshells to be carried out. A triple layer bear-ing of
lead, copper and tin (Glyco 40),together with a bronze bearing with
a
Figure 8: Small bearing test bench 1: Motor; 2: Test bench
housing; 3: Hydraulic cylinder; 4: Endoscope-camera
Figure 9: Large bearing test bench 1: Motor; 2: torque measuring
shaft; 3: Test bench housing; 4: Support bearing;5: Window for
opacity observation
11.2.1 Test Bench and Measurement Method
-
DEVELOPMENT Large Marine Diesel Engines
10 MTZ 12/2001 Jahrgang 62
diameter of 61.5 mm and width of 9.5 mmwere employed. On the
large bearing testbench with the same basic set-up, onlycomplete
plain bearing shells with a diam-eter of 118 mm and a width of 18
mm wereused. Apart from the measuring variablesrequired for
describing the operating con-ditions, such as rotational speed,
drivingtorque, loading force, bearing temperatureand oil intake
temperature, the opacitywas used as a measure of the oil mist
for-mation.
Based on the test bench volume con-taining the oil mist, the
opacity measuredin %OP [10] facilitates a calculation of
thevaporized oil quantity.
Furthermore, the BEAROMOS signalwas measured on the test bench.
BEARO-MOS is a monitoring system developed bySchaller Automation
for bearings and slid-ing surfaces, and is based on the
thermo-couple effect. The shaft and bearing arethus used as a
thermocouple pair. In thecase of dry friction within a bearing,
thethermoelectric voltage that is generateddrives a signal current
through the circuit,which is closed by means of a collector
[11].
The position of the temperature sen-sors on the back of the
bearing can be seenin Figure 10. During the experiments,
ther-mographic, Title Figure, and video record-ings were made in
order to document theoil mist formation.
11.2.2 Tests on the SmallBearing Test BenchThe formation of the
oil mist was demon-strated using half plain bearing shellsand
complete plain bearing shells. Thetest bench was driven at a speed
of 3000rpm in order to simulate the performanceof large diesel
engines as far as possible.The sliding speed is then of the
sameorder as that found in actual largeengines. The diagram in
Figure 11 showsthe measurement results of a test with astatically
loaded bronze half shell. Thestart of the seizure is clearly
concludedfrom the rapid rise at 65 s of the measure-ment time. The
bearing temperatureincreased within a short time from 63 Cto almost
300 C. The oil mist formationcould be observed from a bearing
temper-ature of 200 C onwards. The opacitytherefore increased from
approximately 2% to 75 %.
The oil mist formation began almost 7seconds after the beginning
of the bear-ing seizure. BEAROMOS, on the otherhand, showed a
signal increase evenbefore the torque increase, which indicat-ed
the developing bearing damage. Asthe seizure began, the BEAROMOS
signalonce again increased notably. During the
evaluation conducted after the test, thebearing showed clear
traces of seizureand could have been classified as dam-aged.
A total of about 60 seizure tests wereperformed on the small
bearing testbench. The test results showed that theformation of oil
mist could be stronglypromoted by an increase in the oil
tem-perature. This already proves that the for-mation of oil mist
is very dependent onthe thermal boundary conditions. Theaverage
value of the bearing temperatureat which oil mist formation could
beobserved is around 170 C with a largestandard deviation of 40
%.
The diagram in Figure 12 shows thevalues of the gradient of
opacity relatedto the friction determined in the testbearing. The
almost linear relationshipobserved between the two quantities
isindicated by the interpolation line. Thismeans that a stronger
oil mist formationcan be expected at a higher friction valuein the
bearing. If one forms the quotientof the opacity gradient and the
friction,which is a measure of the gradient of theline in Figure
12, one gets an averagevalue of 0.24 %/(kWs) with a
standarddeviation of 0.08 %/(kWs). It appears
therefore that it is possible to relate theopacity gradient to
the friction, and thusto the occurrence of bearing damage. Itcan be
further concluded from Figure 12that a minimum amount of friction
needsto be induced into the system for oil mistto form. Up to a
certain limit, the develop-ing frictional heat can be dissipated
com-pletely by heat conduction in the bearing,housing and shaft, as
well as by radiationand by the lubricating oil. Only whenthese
possibilities for heat dissipation areno longer sufficient does the
oil heat upstrongly and is consequently subjected tovaporization,
thus resulting in oil mistformation due to drop condensation.
Ananalogue relationship between the fric-tion and temperature
gradient is evident.This is comprehensible, since a higheramount of
friction induces more thermalenergy into the system over time
(heatflow), and therefore the temperatureincrease is also greater.
In the test withcomplete plain bearing shells, the aver-age opacity
gradients were distinctlylower. The average opacity was 0.8 %/s
inthe half bearing shells, while the gradi-ents for complete
bearing shells werebetween 0.01 and 0.25 %/s. The averagevalue for
all the measurements was 0.1
11.2.1 Test Bench and Measurement Method
Figure 10: Section drawing of the small bearing test bench
1: Hydraulic cylinder; 2: Test bearing; 3: Shaft; 4: Support
bearing; 5: Test bearing carrier;6: Temperature test point T1 on
the bearing rear in direction of force application;7: Temperature
test point T2 in highly loaded area
-
11MTZ 12/2001 Jahrgang 62
%/s. The oil mist formation in a completeshell was therefore
slower by almost oneorder of magnitude than in the tests withhalf
shells. This deviating behaviour canbe explained by the better
cooling of theshaft in the case of a complete shell. Thelubricating
oil in the complete shell ismade to flow onto the lower bearing
side,and, after circulation, some of the oilremains in the
lubrication gap, unlike halfbearing shells, in which the
lubricating oilis thrown out. Thus, more heat energy canbe
dissipated from the complete bearingand the shaft. In this context,
the signifi-
cance of the energy balance in the bearingfor the formation of
the oil mist and forthe time progression in the case of oil
mistformation must be emphasized. As can beobserved, changes in the
thermal bound-ary conditions in this case, heat dissipa-tion have a
critical influence on the for-mation of oil mist.
11.2.3 Tests on the Large Test BenchAt present, tests are being
conducted on thelarge bearing test bench with the aim ofverifying
the results obtained from the
small bearing test bench for a larger shaftdiameter. The results
determined so far arepresented in the following. Apart from
thequantities such as loading, torque, bearingtemperature and
opacity described already,the splash-oil temperature, which is
thetemperature of the lubricating oil thrownout from the test
bearing, is also measured.An example of a measurement is shown
inFigure 13. A complete bronze bearing shell,which was loaded by a
force of 3.8 kN, wasused as a test bearing. The bearing seizurecan
again clearly be seen from the torqueincrease, as well as from the
rise in the
11.2.2 Tests on the Small Bearing Test Bench
Figure 12: Gradient ofopacity vs. friction performance for
measurements on thesmall test bench
Figure 11: Measurementon the small test benchunder static
loading with abronze-half plain-bearingshell loading force 3,7
kN,average bearing pressure6,3 N/mm2, Oil inlet temperature 40
C
-
DEVELOPMENT Large Marine Diesel Engines
12 MTZ 12/2001 Jahrgang 62
BEAROMOS signal a short time before. Thebearing temperature in
the highly loadedarea (T2) increases rapidly to over 400 C.An
increase in the splash-oil temperaturecan be observed at about 6
seconds afterthe start of the seizure. By the end of themeasurement
time, the value increasesslowly from 115 C to about 215 C. Oil
mistformation begins about 26 s after the startof the seizure. The
average gradient of opac-ity, which initially was around 0.25
%/s,increases to 1.6 %/s after the second torquerise. During the
measurement, the opacityincreases from 7 % to a maximum of 45
%.This corresponds to about 1 g of vaporizedoil quantity.
Figure 14 shows the relationshipbetween the friction and the
gradient ofopacity, which is already known for thesame bearing test
bench from Figure 12.With the measurements available so far,this
also yields a linear relationshipbetween the two quantities. It is
con-firmed that oil mist is formed only after aminimum amount of
friction. The value ofabout 6 kW, at which oil mist begins toform,
is about one order of magnitudegreater than the corresponding value
ofthe small bearing test bench. It can beassumed that this is the
result of the dif-ferent thermal boundary conditions. Thetest
bearing carrier of the large test benchclearly exhibits a larger
thermal massthan that of the small test bench.
11.2.4 Conclusions from theRadial Bearing TestsOil mist
formation in bearings wasdemonstrated on two test benches
withdifferent shaft diameters. An approxi-mately linear
relationship was found toexist between the induced friction and
thegradient of opacity. It showed that theenergy balance is of
significant impor-tance for the formation of oil mist in
thebearing. The temperature of the suppliedlubricating oil and
other thermal bound-ary conditions therefore have a
decisiveinfluence. The heat energy induced intothe system by the
seizure is completelyremoved by the bearing carrier, shaft andoil.
Only when these facilities for heat dis-sipation do not suffice
does the oil in thegap heat up to such an extent that vapor-ization
occurs. Due to the subsequent re-condensation outside the bearing,
oil mistis finally formed. Damage to the bearingoccurred in all
tests where oil mist wasformed.
However, bearing damage was alsoobserved in bearings where oil
mist wasnot formed. In those tests, the resultingenergy quantities
did not suffice toincrease the gap temperature to such anextent
that the vaporization temperaturerange of the oil was attained.
Outside theinstalled measurement devices for moni-toring the
bearing, the BEAROMOS signalalways responded before an increase
in
bearing temperature or opacity. Theincrease in the splash-oil
temperature fol-lowed that of the bearing temperaturewith little
delay, though with a clearlylower increase rate.
With reference to the OMDEA project, itwas only possible to
outline a summary ofthe experimental results within this arti-cle.
Further results can be found in [12].
11.3 Sliding Block ExperimentsSeveral experiments were performed
inwhich the sliding speed between the slid-ing block and the disc
was altered in orderto adapt to the mean piston speed andpower
output of relevant engine types forwhich OMD monitoring is
considered.
11.3.1 Test Bench and Measuring Technique for Sliding Block
ExperimentsThe sensor arrangement was basicallyidentical to the one
utilized for the radialbearing experiments (11.2). The oil
mistopacity originating in the volume of theenclosed measuring
chamber can also beconverted into an evaporated oil quantity.
11.3.2 Conclusions from theSliding Block TestsThe diagrams
showing the curves of themeasured values, temperatures, opacityand
BEAROMOS signal are comparablewith those resulting from the radial
bear-
11.2.3 Tests on the Large Test Bench
Figure 13: Measurement on thelarge test bench under
staticloading
-
13MTZ 12/2001 Jahrgang 62
ing tests. Therefore, a detailed descriptionis not required
here.
A significant point is the fact that,when a steel disc and a
cast iron slidingblock are paired under the same testingconditions,
the oil mist formation is verymuch faster than with the combination
ofa steel disc and a sliding block of multi-layer material.
11.4 Oil Mist Expansion and OilMist Behaviour in Crankcasesof
Large EnginesSchaller Automation conducted experi-ments on this
subject more than 35 yearsago [13]. However, compared to the
testsperformed in the past, the experimentswithin the framework of
the OMDEA pro-ject require an oil mist generator that per-mits the
evaporation of an amount of oiland its re-condensation into oil
mist(11.2.3), as in the case of real damage.Therefore, the tests
represent an appro-priate simulation of the damage thatoccurs in
engine types monitored byOMD.
An additional target of the investiga-tions is to collate the
tribological experi-ments (11.2 and 11.3) and analyses fromOMDEA
and transform them into dimen-sions of real engine types, which are
mon-itored in order to define Standard Dam-age in relation to oil
mist formation.Moreover, the similarity observations [14]utilized
by Prof. Groth can be of valuableassistance, and the relationships
shown inFigure 12 and 14 are based on these.
11.4.1 The Oil Mist GeneratorA controllable oil mist generator
is essen-tial for the production of the type of oilmist found in
the case of damage. It is aprerequisite that the oil mist achieves
aspecific opacity as formed during realdamage, depending on the
enginecrankcase size, expansion rate and wash-out effect caused by
intensive lubricatingoil splashing. In order to achieve this,
aphysical effect is applied, Title Figure.Lubricating oil under
pressure is heatedup and, as a result of the heat
content(enthalpy), it vaporizes as it expandsunder atmospheric
pressure. In this way,the equivalent oil mist for the simulationof
Standard Damage can be generated.
11.4.2 Experiments on Oil MistSimulation in CrankcasesWithin the
scope of preparing theOMDEA certificate, oil mist
simulationexperiments were conducted with a four-stroke trunk
piston engine with a nomi-nal output of approximately 7000 kWand a
speed of 500 rpm , and a secondfour-stroke engine with a nominal
outputof approximately 8000 kW and a speed of1200 rpm , utilizing a
new oil mist gener-ator.
Figure 16 shows the induction curvesof the oil mist and its
dispersion in thecompartments of a newly developedfour-stroke
engine with a high poweroutput. Analogue results were confirmedand
described by Schaller Automation 25years ago [13]. It should be
noted that
higher thermal loads in modern four-stroke engines also
influence the wash-out effect of oil mist (curves B1 to
B3),resulting from higher splash-oil quanti-ties discharged from
the piston coolingoutlet. The wash-out effect caused by theintense
amount of oil spray is also thereason why it takes longer to obtain
mea-surable values of oil mist opacity in thecompartments located
further away fromthe one in which damage is being simu-lated. The
curves A1 to A3 show the earlyphase of induction per time unit,
contain-ing relatively low amounts of oil mist. Itcan be seen that
both the engine speedand the load have a clear influence on
thedispersion of oil mist along the differentcompartments.
Curves B1 to B3 clearly display thewash-out effect caused by the
oil spray,and the dissipation of oil mist moving inthe compartments
in the direction of thecrankcase ventilation. The
simulationexperiments performed on the high-speed engine show
analogous results.However, the characteristic valuesdeduced from
the B curves point to aslightly higher dissipation factor. It canbe
clearly deduced from both engine cat-egories that efficient engine
protection inthe sense of OMDEA is worthwhile andpossible to
achieve.
12 Summary
The situations described in Parts 1 and 2of this article, and
the results from the
Figure 14: Gradient of opacityvs. friction performance for the
measurements on the large testbench
-
DEVELOPMENT Large Marine Diesel Engines
14 MTZ 12/2001 Jahrgang 62
experimental investigations, show thatpurposeful effort can
indeed result in themeasures required in order to extensivelylimit
the latent operational danger oflarge diesel engines installed in
ships andland-based power stations. Even whenused alone, the
measures introducedwith the OMDEA project will
constitutesubstantial progress. The symptomaticoil mist dispersion
arising from damage,as depicted in Figure 16, A1 to A3, willenable
damage recognition without falsealarms, using intelligent
software.
The inclusion of the BEAROMOS sensorprovides a correlative
sensor systemwhose intelligent evaluation enablesvery efficient
engine protection to be
implemented at a relatively low procure-ment cost.
The use of an additional sensor systemcalled IGMOS (Ignition
Monitoring Sys-tem), which was developed by SchallerAutomation,
will enable the continuousmonitoring of the pressure in the
com-bustion chamber. The protection efficien-cy is maximised by the
comprehensivesystem called DIEMOS (Diesel EngineMonitoring and
Security), which is cur-rently in the final stages of developmentat
Schaller Automation. Not only doesthis system represent
state-of-the-arttechnology with a high ability to provideengine
protection, its procurement cost isdrastically lower than the
extreme costs
encountered when crankcase damageoccurs as a result of
unsuitable devices, inaddition to the fact that the numeroussensors
currently required for traditionalcontrol and measuring are no
longerrequired.
It is hoped that the efforts to collectnew knowledge and the
opportunitiespresented in this article can be of benefitto the
specialists involved, and that thenecessary acceptance of an
effectiveimplementation of engine protectionmeasures to minimize
the risk of thelatent operational danger of large dieselengines can
be improved.
The authors would like to thank allthose who supported their
efforts, and
11.4.2 Experiments on Oil Mist Simulation in Crankcases
Figure 15: Test bench for sliding blockexperiments 1: Disc 300
mm 2: Sliding block3: Drive wheel4: Lubricating oil supply5:
Hydraulic cylinder for axial force6: Opacity sensorThe plexiglass
cover for the measuringchamber has been removed
Figure 16: Oil mist induction A, Oil mist dissipation B for the
consecutive compartments 1 to 9 of the medium speed engine
(Opacity: % OP)A1/B1 for idling speedA2/B2 for 500 1 /min without
loadA3/B3 fr 500 1 /min with load
-
15MTZ 12/2001 Jahrgang 62
they would feel rewarded if this articlewere discussed, with
critical emphasis, inthe Internet Forum at
www.dieselsecuri-ty.org
References
[1] Motor Ship, July 1948, Seite 152ff[2] Burgoyne, J.H.;
Newitt, D.M.: Crankcase
Explosions in Marine Engines, MarineEngineers, 1955, S.
265270
[3] Freeton, H.G.; Roberts, J.D.; Thomas, A.:Crankcase
Explosions, An Investigation intosome factors governing the
selection of pro-tective devices, Institute of Mechanical
Engineers, 1956
[4] Mansfield, W.P.: Crankcase Explosions, Devel-opment of new
protective devices, Instituteof Mechanical Engineers, 1956
[5] Langholz, Jens: Klassifikationsgesellschaftenim Schiffsbau,
S. 36. EuropischeHochschulschriften, Peter Lang,Frankfurt/Main,
1999
[6] SAB-lnebel-Vademekum, S. 31,Abschn.7.2.4. Selbstverlag der
SchallerAutomation Industrielle AutomationstechnikKG, Blieskastel,
1996
[7] Schaller, W.; Drr, M. :Manahmen zurBegrenzung der latenten
Betriebsgefahr vongroen Dieselmotoren auf Schiffen.
MTZMotortechnische Zeitschrift 62 (2001), Nr. 7/8, S. 558 bis
563
[8] Rules and Regulation for Classification ofShips, London:
2001, Control EngineeringSystem, Part 6, Chapter 1, Section 3.1,
Note 4
[9] Uebel, H.: 23. CIMAC Word Congress 2001,Hamburg, Papers Vol.
4, S. 1457oder Homepage der Schaller
Automationhttp.//www.schaller.de
[10] SAB-lnebel-Vademekum, S. 31 bis 33,Abschn. 7.2.4.
Selbstverlag der SchallerAutomation Industrielle
AutomationstechnikKG, Blieskastel, 1996
[11] Schaller Automation: CD ROM DIEMOS(Diesel Engine Monitoring
and Security) kannkostenlos angefordert werden
unter:web:http://www.schaller.de / eMail:[email protected]
[12] Burger, W.; Fritz, M.; Albers, A.: lnebel-entstehung in
Gleitlagern experimentelleUntersuchungen. Tribologie-Fachtagung
2001,GfT und DGMK, Gttingen 2001
[13] SAB-lnebel-Vademekum, S.42 bis 66,Abschn. 7.5 und 7.6.
Selbstverlag der SchallerAutomation Industrielle
AutomationstechnikKG, Blieskastel, 1996
[14] Groth, K.: Geometric Mechanic Similarity, Vorl. Umdruck WS
1985, Universitt Han-nover, Vortrag 1993 in Tokyo
The authors
Dipl.-Ing. Werner SchallerSenior Chief and Mana-ging Director
SCHALLER AUTOMATIONIndustrielle Automations-technik KGBlieskastel /
Saarland /Germany
Dipl.-Ing. Manfred DrrPrincipal Responsible forProject OMDEA at
SCHALLER AUTOMATION
o. Prof. Dr.-Ing. Dr. h.c.Albert AlbersFull professorship at
Insti-tute for Machine Designand Vehicle Constructionof the
University of Karls-ruhe (TH)
Dr.-Ing. Wolfgang BurgerHead of Group Mechatro-nics and
Measuring Tech-niques of Institute forMachine Design andVehicle
Construction ofthe University of Karlsruhe(TH)
Dipl.-Ing. Martin FritzCollaborator of the GroupMechatronics and
Measu-ring Techniques of theInstitute for MachineDesign and Vehicle
Con-struction of the Universityof Karlsruhe (TH)
Univ.-Prof. em. Prof.h.c.mult. Dr.-Ing. habil.Klaus GrothFormer
Head of the Insti-tute for Piston Machinesat the Technical
UniversityHannover / Germany
Acknowledgement
The authors would like to
thank Prof. Klaus Groth for
his excellent advice and
support.
-
DIEMOS
DIESEL ENGINEMONITORING and SECURITY
D - 66432 Blieskastel . Saarland . GermanyTel. + 49 (0) 6842 /
508 - 0 Fax + 49 (0) 6842 / 508 - 260
Die Kooperation mit renommiertendeutschen Forschungsinstituten
unddie Erkenntnisse aus von SCHALLERinitiierten wissenschaftlichen
Projektendienen der Entwicklung von Produkten,die eine
fehlalarmfreie Sicherheitsber-wachung garantieren.
Von Einzelgerten ber Samplingsystemebis hin zu motorspezifisch
angepasstenSicherheitssystemen bietet SCHALLERHardware, Software
und Dienstleistungen,die den sicheren Betrieb groer Diesel-motoren
garantieren.
Unter der Webadressewww.dieselsafety.orguntersttzt SCHALLER eine
non-profit-Website, die als hersteller-unabhngigesForum hierfr
dient.
Die in dem vorliegenden Sonderdruckbesonders hervorgehobene
lnebelde-tektion bildet die Basis der Sicherheits-berwachung, zeigt
aber nur einen Aspektder von SCHALLER AUTOMATION ange-botenen
kognitiven und prventivenSicherheitseinrichtungen.
Ein weiterer wesentlicher Aspekt derumfassenden
Sicherheitsberwachungist die Frherkennung von
Schmiermangel-Schden.
Besonders zu erwhnen ist auch die imRahmen von DIEMOS
angeboteneberwachung der Operating Media.
Somit wird durch die konsequente An-wendung der Komponenten,
Systeme undDienstleistungen, die fr groe Diesel-motoren von
SCHALLER AUTOMATIONangeboten werden ein
Motor-Rundumschutzgewhrleistet, dessen Funktionalitt
undBetriebssicherheit unabhngig von allenanderen
Sicherheitseinrichtungen sind.
Durch den computergesttzen weltweitenReplacement Service RPS ist
die weltweiteVerfgbarkeit der Systeme garantiert.
Sicherheitsbewutsein und Sensibilittim Umgang mit potentiell
gefhrlichenTechnologien nehmen an Bedeutung zu.
SCHALLER AUTOMATION engagiertsich seit ber 40 Jahren fr den
Schutzvon groen Dieselmotoren vor Selbst-zerstrung.
Ausfhrliche Informationen finden Sieauch auf unserer
Homepage:www.schaller.deOder fordern Sie unsere CD ROM an
unter:
[email protected]
Helfen Sie mit, den Dieselbetrieb aufSchiffen sicherer zu
machen. Sagen Sie IhreMeinung und bringen Sie Ihre Erfahrung
ein!
Safety conscience, together with awareness andresponsibility in
managing potentially dangeroustechnologies are taking on more
significance.
For more than 40 years SCHALLER AUTOMA-TION has been engaged in
the protection oflarge diesel engines against severe damageand
their self-destruction.
The cooperation with renowned German rese-arch institutes, and
the findings in scientific pro-jects initiated by SCHALLER
AUTOMATION, isvaluable toward the development of productsthat
warrant a false-alarm-free safety monito-ring.
From single devices utilizing sampling systemsthrough to engine
specific adapted safetysystems, SCHALLER offers hardware as well
assoftware and services that guarantee the safeoperation of large
diesel engines.
The oil mist detection described in this specialedition article
forms the basis of the safetymonitoring; however it shows merely
one aspectof the cognitive and preventive safety equipmentoffered
by SCHALLER AUTOMATION.
Another essential aspect of the comprehensivesafety monitoring
is the early recognition ofdamage resulting from deficient
lubrication.
Within the frame of DIEMOS, emphasis shouldbe applied to the
monitoring of the operatingmedia.
Therefore, through the consequent applicationof components,
systems and services which areoffered by SCHALLER AUTOMATION, a
round-the-engine protection is ascertained, wherebytheir
functionality and safe operation is indepen-dent from all other
manufacturers supplyingsafety equipment.
By means of the computer supported Replace-ment Service RPS, the
availability of thesystems is guaranteed.
Detailed Information is found in our
homepage:www.schaller.de
Or request our CD ROM under:[email protected]
Help us to make the diesel operation safer.Manifest your opinion
by conveying yourexperience!
Under the web pageWWW.dieselsecurity.orgSCHALLER AUTOMATION
supports a non-profit website serving as a manufacturersindependent
forum.