Você está na página 1de 3

THERMAL

EFFECTS

IN

THE

WELD

ZONE

IN

EXPLOSIVE

WELDING

I.

D.

Zakharenko

and

T.

M.

Sobolenko

UDC

621.791.7

In

[1], a description

was

given of a method of measuring

the temperature

in the weld zone,

and cer-

tain assumptions were made that enabled the amount of heat released in the joint to be determined. An estimate of the width of the zone in which the heat is released [1] showed that it is much less than the ampli- tude of the explosive-welding wave.

The

present

study confirms

the

assumption

of [1] concerning the possibility of calculating the thermal

regime in a weld, in which the actual heat source in the weld surface is replaced by an instantaneous source.

Empirical formulas that can be used to calculate the thermal regime of the weld in explosive welding with an accuracy sufficient for practical purposes are also presented, and the calculations are checked by a metallogr aphic method.

Since the

energy QI is released

in

a small

neighborhood of the contact point, it is interesting

to

analyze the possible dependence of Q1 on the impact parameters. In [2] the hydrodynamic problem of the collision of jets of compressible fluid was analyzed in the acoustic approximation: It was found that the

linear dimension of the high-pressure zone is determined by the radius of curvature of the free surface

R =

2

~

y

/

~-

u 2

~

2~1 ~

~+~

sin~ .!_,

2

where

62 is the thickness

U

is

the

flow velocity; c 0 is the

of the lower jet;

y

speed of sound in the jet material; is the collision angle.

form

The

impact

parameters

(y and the plate thicknesses)

are

present

as in the familiar

wave equation,

which for 52 = ~o can be written

= 266~ sin2_L.

2

61 is the thickness

of the upper

(1)

jet;

in this

equation in exactly the same

1971.

TABLE

1

Novosibirsk.

Impact

conditions

h,l~

melt

cale. I tIexpt,

13

7,3

8

17,1

20

21

2,6

1,8

14

9,5

49

40

3

2,9

recrystalliza-

tion

~aic.

expt.

35

3o

33

Material

St.3 +

Cu §

Cu

St. 3

D16 + D16

VQ, ,

km/

Isec

2,25

2,2

1,86

1,86

3.,2

o

o~

Y

18~

16~

'

21~30'

17~

14~20'

13"

28~

21~

3

3

2,5

1,4

2,5

4

6

1

1

10

10

6

2,5

2,5

l0

l0

1

1

8,0

4,3

2,7

2,4

3,55

2,9

3,0

1,44

3,2

5,0

2,9

1,4

1,3

2,4

2,9

2,4

1.0

3,5

 

3,4

Stainless + Stainless

4,5

,"4"i -}-Ti

4

Translated

from

Fizika

Goreniya i Vzryva,

Vol.

7, No. 3, pp. 433-436,

July-September,

Original

article

submitted January

13,

1971.

9 1974 Consultants Bureau, a division of Plenum Publishing Corporation, 227 West 17th Street, New York, N. Y. 10011. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission of the publisher. A copy of this article is available from the publisher for $15.00.

373

!

 

l

2

'~1

GO0

%

 

b

~a

t,

b~$ee

From this we conclude that the angle T and the thicknesses 51 and 52 enter into the relation between Ql and the impact

parameters in the same way as they enter into the expres- sion for R. Moreover, the energy in this zone is naturally

determined by the quantity pUJ, where p is the density of the material, and Uc is the velocity of the contact point.

Thus, we can construct an empirical formula for the energy released in the weld

Ql=kV;Pl6*

~ + ~ ~

sin~

2

"

(2)

Knowing the amount of heat released Ql, for the conditions described in [1] we obtain the value k = 6.8 910-2. In analyz- ing the thermal regime of the welded plates, it is necessary to keep in mind that, in addition to the heat Q1 released in a narrow zone at the interface, heat is also released over the

entire thickness of the material as a result of deformation

[3,4].

Fig. 1, Temperature as a function of time

at various sections in steel for Q1 = 5 cal/

cm 2,T 0=100~

lO-a; 3) 2 910-a.

into heat and uniformly distributed over the thickness of the specimen. In this case the energy released in the weld can be neglected, since it amounts to approximately

3% of the kinetic energy of the projectile plate. Then, for the amount of heat released per unit area of the

plate Q2 we can write

of the specimen

after explosive welding, we assume that the difference be- tween the kinetic energy of the projectile plate and the kinetic energy of the welded plates is completely converted

In order to estimate the temperature

x, cm:

1) 10-3;2)

1.4

9

Q2 = -~" 1 V~O,61(

,2~2

pi Bi + P2B2

).

(3)

As a result of the collision of the plates, the specimen temperature increases by an amount

A T=

Q2

ci p181 ~- c2p~82

where c is the specific heat of the material. The heat ly with the release of Ql; accordingly, in analyzing the tial temperature of the specimen

Q2 is released at the moment of impact simultaneous- thermal regime it is necessary to calculate the ini-

To=TI+AT

,

(4)

where T1 is the temperature before impact.

Thus, we have all the data needed to calculate the temperature in any section of the weld from the ex- pression given in [1]

x 9

 

T=To~

Q~ 2c pV~= a~ t

,e

~,a~t,

(5)

where T o and Q1 are calculated from Eqs. the weld.

(4) and

(2), a

is the thermal diffusivity, and x is the distance from

In order to find the maximum temperature in a section x during the cooling of the Weld, we seek the

extremum of function (5).

an expression for Tmax in the section x

From the condition 0T/0t = 0 we

obtain t = x2/2a 2 .

Tm~x

To ~

Q1

2cpV x

l

e- -E

374

Substituting in

(5), we obtain

(6)

The curves in the figure, calculated from Eq.

(5), show the temperature

as a function of time for steel.

On the wave surface in the eddy zone in thin metallographic sections, and sometimes along the entire im- pact surface, it is possible to observe structures usually caused by preliminary melting of the metal. More- over, recrystallization or relaxation processes and possibly certain other effects can be detected around

the melted areas and near the weld on a distance of 5-50 ~. We analyzed certain impact regimes involving pairs of homogeneous copper-copper, steel-steel, Dural-Dural, titanium-tita~um, and Stainless steel- stainless steel plates.

Using photographs of thin sections cut from plates welded under known conditions, we measured the total melt area along the entire length of the recorded region. For this purpose we employed several photo- graphs of different parts of the same welded plate (from 3 to 8). We then calculated the average thickness h that the melted layer would have had if it had been uniformly present over the entire impact surface and divided in two by that surface. Using this average thickness of the layer converted to the molten state dur- ing the welding process, we determined the minimum amount of heat needed to melt such a layer without al- lowance for heat losses

Q

=

c

m

(T1

--

Tin) § ;~m,

where m is the mass specific heat of fusion.

The amount of heat thus calculated was compared with that calculated from Eq. (2). The results of the calculations and the experimental measurements are presented in the table. As may be seen from these results, the amounts of heat determined experimentally and calculated from Eq. (2) are in good agreement. Since Eq. (6) makes it possible to estimate the maximum temperature near the weld zone, we calculated the maximum distance from the impact surface at which the temperature was still high enough for melting (see table). In this case we again found quite good agreement between the calculated and the actual (averaged) thickness of the molten layer, although Eq. (6) does not take phase transitions into account. As previously noted, around the melted areas and often along the entire weld there are zones of initial recrystallization. Here the grains are very fine, but in equilibrium, and measure on average about 5 ~.

of melted metal per

square centimeter

of plate

area; Tm is the melting point; h

is the

The measured

and averaged width of the recrystallization zone was compared with the width of the

zone of existence of the temperature T2, at which the recrystallization process is still theoretically possible

T2 = 0.4Tm. The results of the calculations and experiments also proved to be similar. Certain discrepan- cies between the calculations and the observations of the width of the recrystallization zone can be attributed to the fact that the recrystallizatibn processes require a certain amount of time. The graph of the time de- pendence of the weld temperature presented in [1] also makes it possible to determine the rate of fall of

temperature

ture in the weld

in the weld.

Thus, on the temperature interval

700-350~

the mean rate of fall of the tempera-

AT -- 3.5. ht

]06

deg/sec.

Obviously, at higher temperatures

greater than the cooling rate in any ordinary metal heat-treating process.

cases metal structures

under the same impact conditions this rate will be even higher

and much

This indicates that in certain

not previously encountered may appear in the weld zone.

LITERATURE

CITED

Fiz. Goreniya i Vzryva, 7, No. 2 (1971).

2. S. K. Godunov, A. A. Deribas,

3. K. Godunov, A. A. Deribas,

4. P. Krasnokutskaya et al., Fiz: i Khim. Obrab. Mater.,

5. A. A. Bochvar, Metallography [in Russian], Metallurgizdat

1o

I. D. Zakharenko,

S.

I.

et

al., J.

Comp. Phys., 5, No. 3 (1970).

et al., Fiz. Goreniya i Vzryva, 7, No. 1 (1971).

No. 6 (1969).

(1948).

375