, standard component
or equipment; interchangeability, ability to be replaced
with another component; survivability, ability of the
product to continue to work after the failure of a consid-
ered component. Contextual criteria includes redundancy,
for components existing in multiple equivalent occur-
rences; Competencies, human required to diagnose and
to repair; Toolings, maintenance equipments like keys,
screwdrivers...; Logistics, delivery of spare parts, trans-
portation of maintenance team...; Environment, working
conditions like lighting, temperature...; Delectability,
easiness to detect failure and components concerned with;
Testability, ability for a component or a sub-system to be
tested...; Maneuverability, ability for a component or
sub-system to be handled; Auto diagnostic: ability for a
system to perform self-testing procedures.
3.2 Maintainability Measure
Maintainability, as a characteristic
tors. They are Maintenance times, Maintenance fre-
quency and Maintenance cost. The three factors are de-
pendant on th e fact that the system is operated and main-
tained in accordance with prescribed procedures and re-
sources. From a systematic perspective, maintenance
includes corrective maintenance and preventive mainte-
nance. And from a software perspective, maintenance
includes adaptive maintenance and perfective mainte-
nance.
There a
inability of a product at the design stage. They are
maintainability design checklists, maintainability evalua-
tion using physical mock-ups, maintainability evaluation
using digital mock-ups and virtual reality and maintain-
ability evaluation u sing quantitative approaches.
The measures used in maintainability analysis
ean time to repair, mean active pr eventive maintenan ce
time, and mean active corrective maintenance time,
maximum corrective maintenance time, and mean main-
tenance downtime [7].
3.2.1 Mean Time to Re
Maintenance time to repair (MTTR) is
depends mainly on the product configurations. MTTR
measures the elapsed time required to perform a given
maintenance activity and is subsequently used to calcu-
late system availability and downtime. Exponential, log-
normal, and normal probability distributions can all rep-
resent mean time to repair. The normal distribution is
normally assumed for mechanical or electromechanical
equipment with a remove-and-replace maintenance con-
cept.
()/
mm
mttri ii
ii
TT
(1)
Figure 3. General maintainability design guidelines
Copyright © 2009 SciRes JSEA
Product Maintainability Design Method and Support Tool Based on Feature Model
168
.2.2 Mean Preven ti ve Maintenance inspections,
3Time
Preventive maintenance activities such as
calibrations, and tuning keep equipment at a specified
performance level. The objective of a preventive main-
tenance program is to postpone the point at which the
equipment or any of its components wears out or breaks
down.
()()
k
mpi pti
i
mp k
pti
i
TF
T
F
(2)
Tmp is the mean preventive time. Tmpi is the elapsed time
Corrective Maintenance Time nting
for preventive maintenance task i, for i= 1,2,3 . . . . . k.Fpti
is the frequency of preventive maintenance task i, for i
=1, 2, 3 , . . . ,k. k is the number of preventive mainte-
nance tasks.
3.2.3 Median
Mean corrective time is a composite value represe
the arithmetic average of the individual maintenance cy-
cle times. Maintainability is the ability o f a product to be
maintained. Calculation of the median corrective main-
tenance time depends on the distribution describing time
to repair. The median corrective maintenance time for
lognormal distributed repair time is given by
2
/exp( )TT
med mttr
(3)
2
is the variance around the mean val
log
ive Maintenance Time tential
ue of the natural
arithm of repair time.
3.2.4 Maximum Correct
This measures the time required to complete all po
repair activities up to a given percentage, often the 90th
or 95th percentiles. The maximum corrective mainte-
nance time for lognormal distributed is
antilog(
mcm m
TTk)
 (4)
Tmcm is th e maximum corrective maintenance time. Tm is
the mean of the logarithms of the repair times.
is the
standard deviation of the logarithms of repair ties. k is
equal to 1.28 or 1.65 for the 90th and 95th percentiles.
3.2.5 Mean Mai ntenance Downtime
m
restore equipment
(5)
Tmmd is the mean maintenance down
ad ld
repair, beginning at time t = 0, will be
aintainability function,
t is time.fr(t) is the probability density function of the
repair time.
ipment can be reached for service, replace-
important
of
ility can lead to
. For example, a
a
Accessibility Value
This is the total time needed either to
to a specified performance level or to maintain it at that
level of performance. Thus it includes active corrective
and preventive maintenance times, administrative and
logistic delay times.
mm
T
dmam adld
TTT
time. Tmam is the
mean active maintenance time, or mean time required to
conduct corrective and preventive maintenance related
tasks. T is the administrative delay time. T is the lo-
3.2.6 Maintenance Function
The maintainability functions are used to predict the
probability that a
gistic delay time.
accomplished in a time t. The m
m(t) for any distribution is expressed by
0
()
t
tr
mftdt (6)
3.2.7 Maintenance Accessibili ty Eval u ation
Accessibility is the relative ease with which a part or
piece of equ
ment, or repair. The lack of accessibility is an
maintainability problem and a frequent cause of ineffec-
tive maintenance.
The evaluation may also be performed by assigning to
each maintainability criterion, a numerical value between
0 and 1 using a table, lik e the one listed in Table 1 [3].
3.3 Maintainability Cost Analysis
Maintainability is an important factor in the total cost
equipment. An increase in maintainab
reduction in operation and support costs
more maintainable product lowers maintenance time and
operating costs. Furthermore, more efficient maintenance
means a faster return to operation or service, decreasing
downtime.
The data to be input into a life cycle cost model in-
clude the purchase price of the product, mean time be-
tween failures (MTBF), MTTR, average material cost of
failure, labor cost per preventive maintenance action,
labor cost per corrective maintenance action, installation
costs, training costs, the warranty coverage period cost of
carrying spares in inventory, and shipment forecasts over
the course of the product's useful life [8].Corrective
Maintenance cost estimation model estimates the correc-
tive maintenance labor cost for a piece of equipment. The
annual cost is expressed by
Table 1. Accessibility evaluation
All the parts ae same
area 1
re directly accessible and placed in th
All the parts are directlerent ar-y accessible and placed in diff
eas 0.8
Some parts are not directly accessible, but those parts are
maintenance free 0.6
Some parts are accessible after disassembling a fast disassem-
bling entity (a screw, etc.) 0.4
The majority of the parts is accessible by disassembling one o
more entities 0
Copyright © 2009 SciRes JSEA
Product Maintainability Design Method and Support Tool Based on Feature Model169
Lmttr
(SOH)(C )(T)
CM
C
mtbf
T
SOH represen
)
ts the scheduled operating hours of the
equipment. CL is the maintenance la
Tmtbf is the mean time between failures for the equipm
Tm
e under
Implied in this defi-
The system reliability,
(7
bor cost per hour.
ent.
ttr is the mean time to repair for the equipment.
3.4 System Reliability Analysis
Reliability can be defined as the probability that a system
will perform properly for a specified period of tim
a given set of operating conditions.
nition is a clear-cut criterion for failure.
Let Ri be the reliability of subsystem i and rij be the re-
liability of component, in subsystem j,
1i
jni=1 ,2, …,k. Then
1
() ()
i
n
iij
j
Rt rt
(8)
say R(t), is given by
11
() 11()
i
kn
ii
ij
Rt rt

 

j
ted in par-
allel, with subsystem i consisting of ni
series for i=1,2,y,k. Such a system i
ries–parallel sy
(9)
The system consists of k subsystems connec
components in
s called a se-
stem [9]. Figure 4 shows the diagram of a
series–parallel system [10].
The system mean time to failure (MTTF) can be
derived in the following form:
1
()
11 ...i
mttf l
lik
n


(10)
1
1(1)
l
k
T
4. The Development and Application of
Product Maintainability Desi
Software
A i-
na Many features can be created once and
gn Support
4.1 Product Feature Library
library feature is a frequently used feature, or comb
tion of features.
then save in a library for future use. Product feature is
organized according to product feature classification of
Figure 1.The product feature can be built and saved in
Figure 4. Series–parallel system
design library of Solidworks. The directory structure can
be built in the directory of install_directory\data\design
library. To create a library feature that includes
references, it is needed to dimension the library feature
relative to the base part on which designer create it. Ref-
erences create dimensions used to position the library
feature (*.sldlfp) on the model (*.sldprt). Steam turbine
feature library in the SolidWorks is shown in Figure 5.
4.2 Productn Feature
e
into they are front axis feature(AFA), front seal
Feature Modelling Based o
Library
Industry steam turbine is composed of about tens of
thousands of parts. The rotor, blade and cylinder of steam
turbine work under the condition of high temperature and
impulsion. They have high manufacture precision. The
rotor of steam turbine is the most important, highest pre-
cision and most complicated part in the steam turbin
product. Rotor includes thous and s of dimensions.
For example, steam turbine rotor axis can be divided
five parts,
feature(AFGS), whole blade wheel feature(ABW), back seal
feature(ABGS) and back axis feature(ABA) .It can be rep-
resented in equation (11).
FA FGA
Fr ACA BGA BA
A
AA A
 (11)
All features of steam turbine rotor can be organized in
the design feature library in the SolidWorks. But the first
st
le which is integrated with the
geometry feature. Steam turbine feature libr
SolidWorks is shown in Figure 5.
ep must be done is to classify the features according to
component classification and recognize the feature di-
mension. Some dimension relation can be built in some
equation. The maintenance feature information had been
organized into attribute tabary in the
Figure 5. Steam turbine feature library in the SolidWorks
Copyright © 2009 SciRes JSEA
Product Maintainability Design Method and Support Tool Based on Feature Model
Copyright © 2009 SciRes JSEA
170
Figure 6. The use case diagram of poduct maintainability design support tool
4.3 The Development Process of PMDSTs
PMDSTs is a typical customization software. It can not
be run without 3D CAD software SolidWorks platform.
The development process of product maintainability de-
sign support software ob eys following methodology.
1) Design the function of product maintainability, de-
sign support tool by using use case diagram.
2) Build main framework DLL (dynamic link library)
using SolidWorks COM Add-In Wizard.
3) Build every function module DLL in C++ language.
4) Call every function module DLL in main DLL
framework by usin g Loa dLib rary method.
5) Test software function and performance by using
SolidWorks Add-in interface.
The use case diagr
ign support tool is shown in Figure 6.The PMDSTs in-
ity and
odel is
by
us
maintainability
de
co re is a kind of
Cd in the server computer.
W
The function StartApplication and TerminateApplica-
tion which is used to connect the solidWorks and termi-
nate the PMDST are as follows.
bool CSteamTurbineSysApp::StartApplication(void)
{ // add menus to the active document
AddMenus();
//Add toolbars
AddToolbars();
// create a control item to handle application-level
events
swAppEvents* eventApp = new swAppEvents;
eventApp->OnCreate(m_pSldWorks);
return TRUE;
}
{ if (m_pSldWorks == NULL)
veToolbars();
if (m_pActiveDoc != NULL)
m_pActiveDoc->Release();
r
am of product maintainability de-void CSteamTurbine-
SysApp::TerminateApplication(void)
s
clude feature modeling, design for maintainabil
reliable analysis use case etc. Product feature m
built in the 3D CAD software Solidworks. Design for
maintainability and reliable analysis is carried out based
on the product feature modeling. The part feature infor-
mation and assemble feature relation can be extracted
ing Solidworks API (Application Program Interface)
function.
The system is developed by using Visual C++ 6.0 and
Microsoft SQL Server 2000. Product
sign criteria are stored in the Database. The database is
nnected using ODBC (Open database Connectivity).
The maintainability design support softwa
/S (Client/Server) software. Windows 2000 Server OS
(Operation system) is installe
indows XP OS is installed in Client computer. The
main framework DLL is created into a visual C++ DLL
or visual C++ .NET DLL add-in using the SolidWorks
COM Add-In Wizard included in the SolidWorks API
SDK (Software Development Kit) [11].
return;
// remove all menus
RemoveMenus();
//remove the toolbars
Remo
// release the PropertyManager object
ReleasePage();
LPMODELDOC pModDoc = NULL;
HRESULT res = TheApplica-
tion->GetSWApp()->get_IActiveDoc(&pModDoc );
if( pModDoc == NULL )
TheApplication->m_pActiveDoc = NULL;
int count = m_EventList.GetCount();
for (int i=0; i<count; i++)
{ CObject* headEvent =
m_EventList.GetHead();
delete headEvent;
Product Maintainability Design Method and Support Tool Based on Feature Model171
in main DLL
fr ::LoadLibrary method is as follows.
tyAnalysis ()
nalysis interface function
AppPath=GetAppPath();
CADServer";
_pSldWorks= TheApplication->GetSWApp();
odDoc = NULL;
m&pModDoc);
OC pPartDoc = NULL;
DD_IPartDoc,(LPVOID
*) I * SeriousDe-
gr SLDWORKS)
;
:\\Steamturbine Main-
s.dll
ca
ocProcAddress(hmod,"
); ULL)
,m_pSldWorks);
4. Index Analysis Case
Throcess block diagram of a steam turbine-generator
.
Mrator
sy
// disconnect from SolidWorks
m_pSldWorks->Release();
m_pSldWorks = NULL;
}
Every function module DLL is called
amework by using
void Maintainabili
{// MaintainabilityA
CString m_AppPath;
m_
CString strConnect;
LPSLDWORKS m_pSldWorks;
// strConnect="MT
m
LPMODELDOC2 pM
HRESULT res =
_pSldWorks->get_IActiveDoc2(
LPPARTD
res = pMod-
oc->QueryInterface(II
&pPartDoc);
typedef void (WINAP
ee)(LPMODELDOC2,LPPARTDO C,LP
HINSTANCE hmod;
hmod = ::LoadLibrary(_T("D
tenanceSoft\\ MaintainabilityAnalysis.dll"));
if(hmod==NULL)
{AfxMessageBox(_T("MaintainabilityAnalysi
n not be found in current directory !") ); }
SeriousDegree lpproc;
lppr = (SeriousDegree)Get
MaintainabilityAnalysis "
if(lpproc!=( MaintainabilityAnalysis)N
(*lpproc)(pModDoc,pPartDoc
FreeLibrary(hmod);
}
4 Maintainability Measure
e p
system is shown in Figure 7 [12]. In the design process
of steam turbine-generator system, MTTR and MTBF etc
aintainability measure index of steam turbine-gene
stem is calculated as follows.
39
()/
mm
i
TT


 227 105.9
370.43
i
mttrii
i
6
10 9
2.69
mtbf
T

96.2
mtbf
T
ATT

mtbf mttr
easure
inde is shown in Fig-
ure 8. Every Sub-system/assembly MTBF value must be
The calculation process of maintainability m
x of steam turbine-generator system
Figure 7. Process block diagram of a steam turbine-gen-
erator system
Figure 8. The calculation of maintainability measure index
ca
5. Conclusions
In this paper, the maintainability design criteria and
measure index used in product maintainability analysis
are discussed. A product feature library for steam turbine
design is built. The feature library can support steam
turbine product feature modeling quickly. Product main-
tainability design method based on feature modeling can
help to analyze product maintainability in the product
design process. PMDSTs is developed by using Visual
tainability design criteria and maintain-
bility measure index in the design stage. This method
will help to enhance product maintainability efficiently.
PMDSTs will be used to product computer support col-
laborative design (CSCD) process through the next step
lculated at first in order to get MTTR, MTBF and
availability etc. index of steam turbine-generator system.
The designer can read Solidworks 3D feature model of
current subsyste m by pressing ‘View 3D Model’ button.
C++ 6.0 and Microsoft SQL Server 2000. PMDSTs can
support designer to evaluate product maintainability by
applying main
a
Copyright © 2009 SciRes JSEA
Product Maintainability Design Method and Support Tool Based on Feature Model
Copyright © 2009 SciRes JSEA
172
of developing new collaborative support function.
6. Acknowledgment
This paper is supported by the Key Technologies R&D
Program of Wuhan City (No. 20081 0321153and Wuhan
Youth Science and Technology Chenguang Pro
(No.200750731289).
REFERENCES
[1] http://www.barringer1.com/jul01prb.htm.
[2] B. Abdullah, M. S. Yusoff, and Z. M. Ripin, “Integration
of design for modularity and design for assembly
hance product maintainability,” Proceddings of 1st Inter-
national Conference 7th AUN/SEED-Net Fieldwise
seminar on Manufacturing and Material processing, pp.
263–269, University Malaya, 2006.
proach for maintainability evaluation in the design proc-
nt (IEEM) 2009.
[6] M. Pecht, “Prainability, support-
ability handboRaton, Florida, pp.
gram 191–192, 1995.
[7]
to en-[9] W. Kue, V. R. Parsad, F. A. Tillman and C. Hwang, “Op-
timal reliability design: Fundamentals and applications,”
Cambridge University Press, 2001.
[10] A. M. Sarhan, “Reliability equivalence factors of a gen-
[3] C. A. Slavila, C. Decreuse, and M. Ferney, “Fuzzy ap-
ess,” Concurrent Engineering, Vol. 13, No. 4, pp.
291–300, 2005.
[4] A. Coulibaly, R. Houssin and B. Mutel, “Maintainability
and safety indicators at design stage for mechanical
products,” Computers in Industry, Vol. 59, No. 5, pp
[12
438–449, 2008.
[5] Y. F. Ding and B.Y. Sheng, “Study on product mainte-
nance integration model,” Submit to The IEEE Interna-
tional Conference on Industrial Engineering and Engi-
neering Manageme
oduct reliability, maint
ok,” CRC Press, Boca
B. S. Dhillon, “Engineering maintainability,” Eelservier,
2008.
[8] H. Reiche, “Life cycle cost in reliability and maintain-
ability of electronic systems,” Computer Science Press,
Potomac, Maryland, pp. 3–23.1980,
eral series–parallel system,” Reliability Engineering and
System Safety, Vol. 94, No. 2, pp.229–236, 2009.
[11] SolidWorks Corporation. “Solidworks 2000 API help,”
2006.
] Stapelberg and R. Frederick, “Handbook of reliability,
availability, maintainability and safety in engineering de-
sign,” Springer, 2009.