Applied Mathematics arid Mechanics Published by SU.

(English Edition. Vol. 16. No. II, Nov. 1995) Shanghai. China

AN EXACT SOLUTION ON THE STRESS ANALYSIS OF FILLET WELDS

Xue D a w e i (~)k:~.t) t

(Received March I, 1995)

Abstract
.Ill t'.X'uct .vohuion on Ihc slre.v.v di.~'lrihuliol! U/'./711cl u't'hl.~ is ohtuincd in this
palWr, This s o h u i o n can he tt.~'cd n o l r m l r /iJr c.~limalin.q, Ihe a c c u r a c r e l the I~rt'.vt'nl
th's(~n method ql'.lilh'l welds hut alstJ.liJr estuhli.~hht~ a new th's&,n methml.

Key words fillet weldment. SilCSS distrihulion, theory ofclasticil)

Abstract
I. Introduction
The one-dimensional problem of the motion of a rigid flying plate under explosive attack has
an analytic
Welding solution only when which
is a technology the polytropic
is xvidehindex
u.,,edofindetonation
the fields products equals toengineering.
of mechanical three. In
general, a numerical
shipbuildi,lg, analysisand
aeronautical is required.
aerospaceInengineering
this paper, and however, by utilizing
bridging. There arcthe several
"weak"dil'lZ'rcm
shock
behavior of the reflection shock in the explosive products, and applying the
kinds of welding such as butt welds, fillet welds und all around welds and can bc distinguished small parameter pur-
terbation method, an analytic, first-order approximate solution is obtained for the problem of flying
into electric arc welding, gas welding, brazing and soldering. I-ligh quality welds can be
plate driven by various high explosives with polytropic indices other than but nearly equal to three.
obtained by both o f atitornalir and protective operations. One Call get a lurge-scalcd, complex-
Final velocities of flying plate obtained agree very well with numerical results by computers. Thus
shaped structure
an analytic formulaorwithelement
two consisting
parameters of of several ntaterials.
high explosive (i.e. So. how io velocity
detonation make a good design on
and polytropic
the strength
index) of weldin,g
for estimation an il'ilporialll
of theis velocity of flyingcnginceriilg i'u'oblcm.
plate is established.
At the prcselll, a lot of devclol~mcnis IlavC boon obtaJllcd in tile fields of x~clding
operations and x~elding materials. People can get ~.i wold ~.vllosc strength is higher thall the
1. Introduction
strength of nlother nliilCl'ia] blii tile rclaling strength design methods of \voiding LII'C very
Explosive
la,,,,in,, driven flying-plate
ill thc.nlonlcnt duc to technique ffmds
Ihc rc;.iSOll thatits toimportant use in thedcsi,-,n
get a \aluabic study nlclhod
of behavior of
is vci"~
materials
complexity under
lind intense
d i f f i c u limpulsive l.lboui sixi.v
t y . Since loading, shock.VCI.II'S
synthesis of diamonds,
ago tip and explosive
tO next,, designers welding
el" x~clcts hilveand
nol
cladding of metals. The method of estimation of flyor velocity and the way
knowll the cxilct distribution rtllc of Stl'CSSCS in wolds, tile only thing ihcv can tit) in tile of raising it are questions
of commondesign
strength interest.
is It) calculate the values of imaginal avel'agc stress uciiilg on ihc Ihroai pUl't of
Under the assumptions of one-dimensional plane detonation and rigid flying plate, the normal
welding Seilnl ilccordhlg tO Ihe ii'aditioilal design melhod. Assi.llllhlg Ihut the rlornl;.l] Ml'CSs is
approach of solving the problem of motion of flyor is to solve the following system of equations
unil'ormly dJstribulcd on the hi.lit ~tclcls is perhaps accepted by the people bcCatl~,C it is rather
governing the flow field of detonation products behind the flyor (Fig. I):
safe (Pig. I). But the su.ggcstlons of tile relating noiTnal slress and siicar slrcss acting on the
t h r o a t section of tile fillet ~velding seam (Pig. 2) alld both of the stresses to he tlscd as the
--ff +u_~_xp+ au =o,
strength control I]ictor of the Seilnt ap are perhaps rio[ i'OilSOlli.lble. So thai. a lot of scientists have
done their best in order to h'nprove au Ihe n'leihod
au of1 welding strength design. In the year of 1945.
y =0,
a stress distribution diiigran'l of welding seilnl wits drilwn by Norris Ill based on photoelastic (i.0
method as shown hi Pig. 4(a) whichaSis widely a s used and published once again in a textbook for
university study 121. But his diagram a--T =o,
is not satisfactory if we make a c o m p a r i s o n on the exact
=p(p, s),
stress distribution suggested in thisp paper for the same problem. Another diagram shown in :
Fig. 5(a) as a part of a review o f literatures before the year of 1937 on the distributions o f
where p, p,stresses
principal S, u areand
pressure, density,
maximum specific
shear stressentropy
acting and particle
on the velocity
throat sectionof Of
detonation products
welditig seam was
respectively, with the trajectory R of reflected shock of detonation wave D as a boundary and the
trajectory F of flyor as another boundary. Both are unknown; the position of R and the state para-
9t The University of Macau
meters on it are governed by the flow field I of central rarefaction wave behind the detonation wave
D and by initial stage of motion of flyor also; the1(119
position of F and the state parameters of products
293
 1020 Xue Dawei
published by Salakian and Claussen TM. This diagram was also collected in the same textbookt-'L
The relating exact diagram given in this paper for the same problem is shown in Fig. 5(b). It is
obviously that these two diagrams shown are far away from each other. Design methods of
welding strength problem issued in other textbooks for university study 14 51 and the well-known
mechanical engineering handbook 16t are just the same as the books written more than fifty
years ago, Thus, we can conclude that there has been no outstanding contribution offered [br
the welding strength analysis problem for such a long time.

Abstract
l .
The one-dimensional problem of the motion of a rigid flying plate under explosive attack has
an analytic solution only when the polytropic index of detonation products equals to three. In
general, a numerical analysis is required. In this paper, however, by utilizing the "weak" shock
behavior of the reflectionFig. shock
1 in the explosive products, and applying theFig. small2 parameter pur-
terbation method, an analytic, first-order approximate solution is obtained for theinproblem
An exact solution on the stresses distribution of fillet welds is suggested this paper. This
of flying
plate driven
solution canbybevarious
used high explosives
not only with polytropic
to estimate indices
the accuracy of other than butdesign
the present nearlymethod
equal toofthree.
fillet
Final
welds velocities
but also ofto flying platea obtained
establish new ~lesignagree very well
method. Wewith numerical
suggest resultscontrol
a strength by computers. Thus
factor whose
an analytic
value equalsformula
to sevenwith twoofparameters
times the value of
of high explosive
average normal (i.e.stress.
detonation velocity and polytropic
index) for estimation of the velocity of flying plate is established.
II. The Derivation of the Exact Solution
1. Introduction
A fillet welding and the relating loading are shown in Fig. 3(a). Let / denote the length of
the seam and k the radius of the cross section of the seam which has the shape of a quarter of
Explosive driven flying-plate technique ffmds its important use in the study of behavior of
a solid circle. Assuming that this is a plane problems of theory of elasticity and take the polar
materials under intense impulsive loading, shock synthesis of diamonds, and explosive welding and
coordinate
cladding (r. tl) asThe
of metals. shown in Fig.
method 3(b)
of estimation of flyor velocity and the way of raising it are questions
The exact
of common interest.solution of such a problem has to satisfy the following conditions:
I.Under the assumptions
The resultant of one-dimensional
of ~'r0 acting plane
on 0 = 0 equals - P detonation
: and rigid flying plate, the normal
approach of solving the problem of motion of flyor is to solve the following system of equations
/t

governing the flow field of detonation oz,o(r,

I o)ldr=-p
products behind the flyor (Fig. I): (2.1)

--ff
ap +u_~_xp+ au
=o,

au au
y1 =0,
(i.0
hi2 aS as
a--T =o,

respectively, with the trajectory (a)

p =p(p, s),

1
where p, p, S, u are pressure, density, specific entropy and particle velocity of detonation products
R of reflected shock of detonation wave D as a boundary and the
(b)
trajectory F of flyor as another boundary. Both are unknown; the position of R and the state para-
meters on it are governed by the flow field I of central rarefaction wave behind the detonation wave
Fig. 3
D and by initial stage of motion of flyor also; the position of F and the state parameters of products
293
 A n Exact Solution on the Stress At~talysis o f Fillet Welds 1021

2. T h e resultant o f o'd acting on 0 = 0 equals zero:

~icr,(r, o)Idr----'O (2.2)

3. T h e resultant o f m o m e n t o f G0 acting on 0 = 0 equals Ph 2:

fooo(r, O)lr dr=_ 2Ph (2.3)

4. T h e resultant o f or0 acting on 0 = r r / 2 equals P:

5. The resultant m o m e n t o f crt acting on 0=rr, 2 equals zero:

Abstract

The one-dimensional problem of the motion of a rigid 2 =0flying plate under explosive attack (2.5)
has
an analytic solution only when the polytropic index of detonation products equals to three. In
general, a numerical
6. The resultant analysis is required.
o f rt0 acting on 0 = nIn 2this paper,
equals however, by utilizing the "weak" shock
zero:
behavior of the reflection shock in the explosive products, and applying the small parameter pur-
terbation method, an analytic, first-order approximate solution is obtained for the problem of flying
plate driven by various high explosivesIir,o(r, 2)ldr=O
with polytropic (2.6)
indices other than but nearly equal to three.
Final velocities of flying plate obtained agree very well with numerical results by computers. Thus
7. The n o r m a l stress tr, (r. 0) at r = h equals zei'o:
an analytic formula with two parameters of high explosive (i.e. detonation velocity and polytropic
index) for estimation of the velocity of flying plate is established.
or, (h, O) = o (2.7)

1. Introduction
8. The shear stress rr0 (r, 0) ;.it r = h equals zero:

Explosive driven flying-plate technique ffmds its important use in the study of behavior of
r,~(h, O)=o (2.8)
materials under intense impulsive loading, shock synthesis of diamonds, and explosive welding and
cladding of metals. The method of estimation of flyor velocity and the way of raising it are questions
9. The interest.
of common shear stress rt8 acting on the cross section o f f l = 0 and the same shear stress
actingUnder
on thethecross-section
assumptions oof one-dimensional
f 0=Tr 2 at point Oplane detonation
a r e equal andother.
to each rigid flying plate, the normal
approach of solving the problem of motion of flyor is to solve the following system of equations
0)---,,,(0,
governing the flow field of detonation products behind the flyor (Fig. I):

10. The equilibrium e q u a t i o n s--ffalong the r and=o, 0 directions respectively: have to be

ap +u_~_xp+ au
satisfied.
I I. The values o f all the c o m p o nau
e n t s o fau
slressy1are not
=0,infinite.
We a s s u m e t h a t (i.0
aS as
ae= (a+br+cr
a--T ~) (1=o, - sin0) +asin0 (2.10)

where, a. b and c are =p(p, s), and cr~--p/lh.

c o n s t a n t to be pdcte,'mined
The equilibrium e q u a t i o n along the 0-direction o f an element is
where p, p, S, u are pressure, density, specific entropy and particle velocity of detonation products
respectively, with the trajectory R of reflected shock of detonation wave D as a boundary and the
1 03 (rZr,e) + 1r 03ae the position of R and the state para-
trajectory F of flyor as another boundary.
r z 03r Both are unknown;
---~----0 (a)
meters on it are governed by the flow field I of central rarefaction wave behind the detonation wave
D and by initial stage of motion of flyor also; the position of F and the state parameters of products
293
 '3'3
I()__ Xue Dawci
Substituting the formula (10) into the above-mentioned equation we have

(9
dr (r2r"g) = ( a r + b r : + c r 3) c o s O - crrcosO (b)

this equation has its general solution its

za br
~,o=~T+T+ -~)cos0- TcosO-I
cr f,(O)
r~ (c)

~hcrc..l;(0) is an arbitrar> function of 0. this function has to b c o m i t t e d due to the reason that
the \alne o f r,.0 can not be infinite. So that we have

l a . br . cr z \ . (r
" t - - ~ " t - - ~ ) C OSU -- V C OSO (2. l l)

Abstract
The equilibrium equation o f an infinitesimal clement along the direction of r is

The one-dimensional problem o~crrof the motion

1 ~rr0 { of a rigid flying plate under explosive attack has
,rr--(ro
=l,
an analytic solution only when Orthe {polytropic ," ,30 indexr of detonation products equals to three. (Ind )
general, a numerical analysis is required. In this paper, however, by utilizing the "weak" shock
Sttbstituting
behavior the f o r mshock
of the reflection u l a e { 2in. 1 the
0 ) a nexplosive
d (2. l l ) i nproducts,
t o the equation
and applyingx~c have
the small parameter pur-
terbation method, an analytic,
__
first-order approximate
I a
solution
2br
is
3or 2
obtained for the problem of flying
plate driven by various high explosives
( rcr, ) = a q - b rwith + cr polytropic
~ - ( - ~ - q - - - -indices
~ nt - -other ~ ) but nearly equal to three.
than
4 g sinO (e)
dr
Final velocities of flying plate obtained agree very well with numerical results by computers. Thus
an
andanalytic formula
then \~c with
can =,,et the two parameters
folloxvint~ of highfrom
expression explosive (i.e. detonation
the general solution ofvelocity and polytropic
this equation
index) for estimation of the velocity of flying plate is established.
_br. cr z l a. t, , cr z or\ /~(0)
o',---- a -t--~--t- ~ - - --- ~--or-t
~, 2 3 4 2-]sinO q
r (f)
1. Introduction
x~here. Z)(0) is an arbitrary function of 0 as x~cll, this function has to bc omitted as the values
o f o ' r Explosive driven
c a n Ilot bc flying-plate
infinite. technique
So that we have ffmds its important use in the study of behavior of
materials under intense impulsive loading, shock synthesis of diamonds, and explosive welding and
cladding of metals. The method
cr,=u.+~ _i.._.~
" cr~ _ of
of estimation ( 2 flyor
+ "31velocity
brat
c rand
4
2 G .)sin0
the
-2 (2.12)
way of raising it are questions
of common interest.
Under the assumptions of one-dimensional plane detonation and rigid flying plate, the normal
The values o f constants a. h. and (' in I\wmulac (2.10). ( ! 1 1 ) and (2.12) can be determined
approach of solving the problem of motion of flyor is to solve the following system of equations
governing the flow field of detonation products behind the flyor (Fig. I):
From condition (2.9) we have

--ff afa =o, (2.t3)

au ap +u_~_xp+
substituting the formula (2.10) into formt, hm(2.2) and (2.3) and paying attention to (2.13) we
au au 1
I'~z|x'C y =0,
(i.0
aS as
a--T~h a-+- -~h c -I =o,
_
- o'
(g)
p =p(p, s),
h ht
3 b +-~-c---- o (h)
where p, p, S, u are pressure, density, specific entropy and particle velocity of detonation products
respectively, with the trajectory R of reflected shock of detonation wave D as a boundary and the
son ing these equations about b and t' simultaneously we have
trajectory F of flyor as another boundary. Both are unknown; the position of R and the state para-
meters on it are governed by the flow field I of 18o" _ rarefaction
central 240" wave behind the detonation wave
b = also;
D and by initial stage of motion of flyor - - - - ~ the
- - , position
, ; = ~ of F and the state parameters of products
(2.14)
293
 An [[xact Solution ola the Stress Analysis of Fillet Welds 1023
Substituting the values o f u, /~, and c into the fo"mulae (2.10)(2.11) and (2.12). we can get
all the expressions o1" all the c o n l p o n e n t s o f stress as follows:

cry.= a l~l r rz r r

(2.16)

-~ )oo 0 (2.17)

It is eusil5 proved that fornlulae (2.15), (2.16) and (2.17) satisl)' all the conditions
expressed by (2.1) to (2.11). In fact. c o n d i t i o n s (2.2), (2'.3) and (2.9) were satisfied when
d c t c r m i n g the c o n s t a n t s o f a. /~ and c: from formula (2.17) and letting 0 = 0 we have
Abstract

The one-dimensional problem of the motion of a rigid flying plate under explosive attack has
(i)
an analytic solution only when the polytropic index of detonation products equals to three. In
and substituting
general, a numericalintoanalysis formulais (2.1) we can
required. In seethis that
paper, however, by utilizing the "weak" shock
behavior of the reflection shock in the explosive products, and applying the small parameter pur-
terbation method, an analytic, first-order approximate solution is obtained for the problem of flying (J)
plate driven by various high explosives with polytropic indices other than but nearly equal to three.
Final
s o thnt velocities
c o n d i tof
i o nflying
(2.1) plate obtained we
is satisfied: agreek nvery
o ~ well that with
a..=rrnumerical
is a c o results
n s t a n t by
at computers.
0=re/2, soThus
that
an analytic formula with two parameters of high explosive (i.e. detonation velocity
c o n d i t i o n (2.4) and (2.5) are all satislied: fiom (2.17) we know that v,.0 (r. ' r t 2 ) = 0 at 0 = r e 2 and polytropic
index)
and then for estimation
c o n d i t i o n of(2.6)
the velocity of flying
is satisfied: the plate
c o n disi testablished.
i o n s (2.7) and (2.8) are also satisfied from
I\~rmulae (2.15) and (2.17) if we take r = h into these two later formulae and the conditions (2.
1. Introduction
10) and (2. I I ) are oF course satisfied because these c o n d i t i o n s were utilized before.
O f course, xve can easily. =,,et the c o r r e s p o n d i n g expressions o f c o m p o n e n t s o f strains and
d i s p l aExplosive driven flying-plate
c c m c n l respectively by usingtechnique ffmdslaw.
the Hooke's. its important use in the
formulae (2.15). study
(2.16) and of(2.17)
behavior
and of
the
materials under intense impulsive loading, shock synthesis of diamonds, and explosive
relations bct\veen strain and displacement. Here. all o f these expressions are onlitted, because i]
welding and
cladding of metals. The method of estimation of flyor velocity and the way of raising it are questions
is not necessary in design. So that. ~ve can conclude that an exact solution on the analysis o f
of common interest.
l]llet Under
welds is theo bassumptions
t a i n e d here.of one-dimensional plane detonation and rigid flying plate, the normal
approach
III. S o m ofe solving
C o m pthe a r i sproblem
oos of motion of flyor is to solve the following system of equations
governing the flow field of detonation products behind the flyor (Fig. I):
The results given by Norris shown in Fig 4(a) and the results given by this p a p e r for the
same p r o b l e m shoxvn in Fig. 4(b) are not the same. After a c o m p a r i s o n between these results
--ff
ap +u_~_xp+ au
=o,
~ e can see that:
I. The shear stress distribution I~1
auon b oau
u n d a r y1 O C is wrong: the exact solution sho~vs that
y =0,
this stress equals zero. (i.0
2. The n o r m a l stress ~r acting aS on b o uansd a r y O C ul is wrong. It is necessary' to emphasize
that the p h e n o m e n o n o f stress c oa--T
n c e n t r a t i o n =o,
at point O in Fig. 4(a) can not happen. IF
s o m e o n e makes a strength design basedp =p(p, s), results shown in Fig. 4(a), then the wrong value
on the
o f n o r m a l stress will function as a control factor.
where3.p,All
p, S,
theucomponents
are pressure, acting
density,
on specific entropy Aand
the boundary O IHparticle velocity of detonation products
are wrong.
respectively, with the trajectory R of reflected shock of detonation
So that, the resultg given by Norris can not be used for a strength wave Ddeisgn
as a boundary and the
of fillet welds.
trajectory F of flyor as another boundary. Both are unknown; the position of R and the state para-
The results introduced from Salakianj and Claussen's review are shown in Fig. 5(a) and
meters on it are governed by the flow field I of central rarefaction wave behind the detonation wave
Dtheandr by initial stage
diagram
of motiongiven by the
of flyor also;exact solutio~
the position of Ffor
andthe
the same problem of
state parameters is shown in
products
Fig. 5(b).
293
 1024 Xue Dawei

(7 B a, B

~s

}' (a) (b)

Fig. 4
Abstract

The one-dimensional problem of the motion of a rigid flying plate under explosive attack has

~
an analytic solution only when the polytropic index of detonation products equals to three. In
general, a numerical analysis is required. In this paper, however, by utilizing the O't
"weak" shock
behavior of the reflection shock in the explosive products, and applying the small parameter pur-
terbation method, an analytic, first-order approximate solution is obtained for the problem of flying
plate driven by various high explosives with polytropic indices other o than but nearly
B equal to three.
O B
Final velocities of flying plate obtained agree very well with numerical results by computers. Thus
an analytic formula with two parameters of high explosive (i.e. detonation velocity and polytropic
index) for estimation of the velocity
(a) of flying plate is established. (b)

Fig. 5
1. Introduction
It is obviously that the dangerous point of the seam is at point O as shown in Fig. 5(a),
but this point isdriven
Explosive not tile dangeroustechnique
flying-plate point as ffmds
shownits inimportant
Fig. 5(b).useThe
in value of stresses
the study at point
of behavior of
materials
O just asunder
shownintense
by theimpulsive loading,
exact solution ~reshock
a r = c rsynthesis
0 = c r , r ,of
~ =diamonds,
0. and explosive welding and
cladding
It isofimportant
metals. Theas method
shown by of estimation of flyor velocity
the exact solution that theand the way
throat of raising
section it are questions
of a welding seam is
of
notcommon
the placeinterest.
where the stress takes its rnaximum value. We suggest that the designers use the
exact Under
formulathe given
assumptions of one-dimensional
by (2.15), (2.16) and (2.17) planewhen detonation
they make and their
rigid strength
flying plate, the normal
design of'fillet
approach of solving the problem of motion of flyor is to solve the following system of equations
~velds. We suggest that the strength control factor be the normal stress of ~r0 at r = h and (i=0
governing the flow field of detonation products behind the flyor (Fig. I):
where the value of stress is seven times of the average stress.

References --ff
ap +u_~_xp+ au
=o,
[1] C. H. Norris. Photoelastic investigation of stress distribution in transferse fillet welds,
au au
Welding Journal, 24 (19"~5), 557. y1 =0,
[2] (i.0
J. E. Shigley and C . R . Mischke, Mechanical Engineering Design, Fifth ed., McGraw-Hill
aS as
Book Co., New York, (1989), a--T
I - 779. =o,
[3] A. G. Salakian and G. E. Claussen, Stress distribution in fillet welds: A review of the
p =p(p, s),
literature, WeMing Journal, 16 (*1937)~ 1 - 2 4 .
[4] M. Faires Virgil, Design of Machine Elements, Fourth ed., The Macmillon Company,
where p, p, S, u are pressure, density, specific entropy and particle velocity of detonation products
New York,
respectively, with the(1965), 624. R of reflected shock of detonation wave D as a boundary and the
trajectory
[5] R. F
trajectory flyor asFundamentals
M.ofPhelan, of Mechani'cal
another boundary. Design, Third
Both are unknown; ed,, McGraw-Hill
the position Bookpara-
of R and the state Co.,
meters New
on it York,
are governed by the flow field I of central rarefaction wave behind the detonation wave
(1970),-1-678.
D[6]
and by
M. initial
Kutz, stage of motion
a n d other of flyor also;
-Mechanical the position
Engineers of F and
Handbook,. the state parameters
John-Wiley and Sons, ofNew
products
York,
(1986i, 1 - 2 3 1 6 . 293