Analytical study on the effects of welding deposition sequence on microstructure and hardness in beam to column welded connection of steel framed structures

It is evident from the figure that 1 layer 1 pass welding at a constant speed was performed for all the welding types, the continuous uni-directional, the continuous turning back and the

For steel framed structures1,2, of the type to which

enormous damage was seen during the Northridge

Earthquake and the Great Hanshin–Awaji Earthquake3,4,

it is essential to have prior understanding of the mechanical

properties, such as the strength and toughness, which

vary depending on the welding conditions, in order to

ensure the welded joint quality

Experimental with the effects of welding conditions

upon joint performance are enthusiastically pursued,

including for steel framed building structures.5–8 Numerical

simulation techniques are thought to be effective tools

to investigate these results in further detail and to relate

the above with design and fabrication These techniques

are also effective tools to further understand the

temperature history during welding and phenomena such

as phase transformation behaviour

Accordingly, this article aims to explicate the effects

of multiple thermal cycles upon the welded joint

performance; in Report 19 an understanding was achieved,

by experiment, of the effects of the welding conditions

upon the joint performance of beam-to-column welded

connections of examples of steel framed structures The

effects of the differences in the welding thermal cycles

upon the experimental results were also investigated by

means of numerical simulation In more detail, by means

of experiment, an investigation was made into the effects

of the welding conditions, such as heat input and interpass

temperature upon the strength and toughness of the

welded joints made by various welding conditions; an

understanding of the joint performance characteristics

was achieved for the actual fabrication conditions

employed for steel framed structures Furthermore, the

temperature history of the entire joint, including the start

and finish zones, and the actual interpass temperature

conditions were analysed in detail by means of

three-dimensional simulation which took the moving heat source

into account These experimental results and the further

consideration by means of numerical analyses reconfirmed

the importance of the temperature distribution control

in order to ensure optimum joint performance

In the study described in this article, the temperature/

microstructure/hardness during the welding of

beam-to-column welded connections of steel framed structures

were simulated; an investigation was carried out on how the joint microstructure and hardness vary due to the welding deposition sequence, in addition to the difference

in the temperature history studied and described in the previous report That is to say, the temperature history, the microstructure and the Vickers hardness of various systematised welding deposition sequences were analytically obtained using a three-dimensional multi-layer welding simulation which took the moving heat source into consideration Next, the relationship between the cooling rate and the hardness at the weld zone was investigated using analytical results with the cooling time t8/5 from 800 oC to 500 oC as a parameter indicating the cooling rate In addition, an investigation was carried out into the effect of the welding deposition sequence upon the HAZ width and the Vickers hardness of the heat affected zone Consideration was also given to an analytical prediction of the welded joint tensile strength

Multi-layer welding of beam-to-column welded connections of steel framed structures

The joint configurations and the welding procedure

conditions were faithfully simulated within the scope of

actual procedure conditions employed in the welding site of steel framed building structures in the production

of weld test specimens and the construction of analytical models in the previous report9; the differences between the analytical and experimental results were compared and studied By contrast, for this report, the welding simulation was performed with an emphasis on the systematic arrangement of the detailed results obtained

by numerical analyses rather than the interpretation of the experimental results; for that reason, the conditions mentioned below were established by selection of the optimum state for the joint configuration and the welding conditions

Test specimens used for analysis

A butt welded joint with a beam flange weld zone appropriate to a beam-to-column welded connection of steel framed structure, as indicated in Fig 1, was selected

as the configuration employed for the numerical analysis

Analytical study on the effects of welding deposition

sequence on microstructure and hardness in

beam-to-column welded connection of steel framed structures

M M O C H I Z U K I, Y M I K A M I and M T O Y O D A

Department of Manufacturing Science, Graduate School of Engineering, Osaka University

The plate widths of the diaphragm, flange and backing

ring were all 200 mm and it was assumed there was no

excess metal since the effect of this upon the weld heat

input and the interpass temperature was thought to be

minimal The material employed was structural 490 MPa

class steel for buildings and an analysis was carried out

in which the carbonic acid gas shielded arc welding

process was simulated The figure also shows the

measuring point of the interpass temperature recommended

by the Japanese Architectural Standard Specification

6-Steel Work (hereafter referred to as JASS 6)10, which is

in the central zone on the axis of the weld line, on the

base metal side 10 mm from the upper end of the groove

Analytical method for microstructure and

hardness

Figure 2 shows the CCT diagram for 490 MPa class steel

used for the assumed cases Four phases of ferrite

(regarded as the same structure as pearlite), bainite,

martensite and austenite were taken into consideration

and, prior to welding, the base metal was assumed to be

ferrite single phase The horizontal axis represents the

cooling time from 900 oC The extended Johnson–Mehl–

Avrami rule11 was employed for the calculation of the

phase proportions during cooling and the microstructural

proportions were calculated from the temperature history/

cooling rate at each position obtained by the thermal

conduction analysis The figure indicates several cooling

curves and the cooling time t8/5 from 800 oC to 500 oC was

also shown as a standard for the cooling rate The

numerical values indicated at the lower side within the domain of each phase are the phase proportions of each phase following cooling when the cooling took place along the cooling curve For example, observation of the cooling curve second from the right shows that the cooling time t8/5 from 800 oC to 500 oC is 150 s and a microstructure

of 44 % ferrite and 56 % bainite is indicated following cooling For materials which are subjected to multiple thermal cycles such as multi-layer welds, the actual material properties undergo change each time the material goes through the heating/cooling process; however, realistically, it is very difficult to obtain the values of the physical properties which represent the above mentioned,

so the investigation proceeded, on this occasion, with the assumption that the material properties which were conveniently given as the initial state were maintained for the purpose of the analysis

Table 1 shows the chemical composition of the 490 MPa class steel used for the calculation of the hardness

of each phase For this study, the Vickers hardness of each phase was calculated using an empirical equation with the chemical composition and the cooling rate VR (K/h) at 700 oC as parameters and the Vickers hardness were estimated by the mixing rule using the phase proportions of each phase The following equations were employed to calculate the Vickers hardness of each phase employed in this study:

HvF = 62 + 223C + 15Si + 30Mn + 21Ni +

HvB = –323 + 185C + 330Si + 153Mn + + 65Ni + 144Cr + 191Mo + log10VR (89 + 53C – 55Si – 22Mn – 10Ni – 20Cr – 33Mo) [2]

HvM = 127 + 949C + 27Si + 11Mn + 8Ni +

Here, the Vickers hardness for each phase of ferrite, bainite and martensite is HvF, HvB and HvM respectively From

these and the phase proportions of each phase pF, pB and pM the Vickers hardness distribution, following welding, was estimated at each position using the mixing rule as in the following:

HvAll=pFHvF + pFHvF + pFHvF [4]

A temperature/microstructure/hardness simulation, taking into account the proportion of the microstructure which changes with time and accompanying microstructural dependence of the physical properties, under these conditions, was carried out using the general-purpose finite element method code SYSWELD14 The details of the analysis, other than that for the

1Configuration of a beam-to-column welded joint specimen.

2 CCT diagram of material used.

7

0 0 3 1 7 0 7 0 6 0 0 8 0 0 2

Table 1 Chemical composition of material used

microstructure and Vickers hardness described above,

are as indicated in the previous reports.15–17

back When the temperature exceeds 350 oC, welding is performed in the same direction as the preceding pass after the appropriate weld dwell time)

To facilitate a rigorous evaluation of the HAZ, a heat input pattern was selected for use such that the whole area of the weld metal zone exceeds the molten temperature with a constant welding speed (5 mm/s) Figure 3 shows

a comparison of the heat input from analysis for each type of weld It was assumed that the weld heat input Q was estimated as the heat value actually imparted and the weld heat input equivalent for the analysis, which takes into account a normal heat transfer, was imparted such that the finite element model was made to have adiabatic boundary conditions; the weld heat input was calculated from the mean temperature rise after the temperature distribution became uniform using the following equation:

Q = (Cp × W × DTav)/L [5] Here, Cp : Specific heat, W : Mass, ∆tav : The mean temperature rise assuming an adiabatic boundary, L : Weld length

Furthermore, the initial layer was called ‘First’, the fourth layer, which was the middle layer, was called ‘Mid’ and the 7th layer which was the final layer was called

‘Last’; the heat input, with a constant welding speed for the entire joint, was investigated by comparing the weld heat input of these three layers It is evident from the figure that 1 layer 1 pass welding at a constant speed was performed for all the welding types, the continuous uni-directional, the continuous turning back and the interpass temperature controlled welding types and thereafter the weld heat input increased with the increased bead width following the superimposed layers For the interpass temperature controlled welding type, where weld dwell was inserted in order to control the interpass temperature of the final layers, the weld heat input to the last layer was increased more than the other types because the whole test specimen was subjected to cooling These results were seen to have a similar heat input pattern

to that seen during the fabrication of the weld test specimens described in the previous report.9

Table 2 Parameters of welding pass sequence

3 Comparison of heat input in multi-pass welded joints.

Welding pass deposition sequence

The welding pass deposition sequence and the welding

sequence were determined for each type as indicated in

Table 2 so that a systematic arrangement of the results

could be performed while referring to the welding

conditions for actual beam-to-column welded connections

of steel framed structures Interpass temperature controlled

welding and continuous welding were employed in a 7

layer 7 pass welding sequence in order to vary the interpass

temperature In addition, the following types of continuous

welding were carried out: continuous uni-directional

welding in which welding is performed constantly from

a fixed direction and continuous turning back welding

where welding is performed by turning back at the weld

start and finish zones The effects of the turning back

operation upon the weld start and finish zones were

investigated in detail:

• Continuous uni-directional welding: Build-up welding

uni-directionally and continuously irrespective of the

interpass temperature (The time taken to start welding

from the finish zone to return to the start zone is

as-sumed to be, ideally, nil and welding of the succeeding

layer is assumed to commence immediately after the

completion of welding of the previous layer.)

• Continuous turning back welding: Continuous

build-up welding, turning back at the end zone irrespective

of the interpass temperature, as employed to some

ex-tent in steel framed building structures (The

succeed-ing layer weld metal is assumed to build-up directly

above the previous layer and immediately after the

com-pletion of welding of the preceding pass.)

• Controlled welding: Build-up welding uni-directionally

with control of the interpass temperature to be less

than 350 oC at a position 10 mm from the groove end

zone of the flange centre This complies with the

con-ditions recommended by JASS 6 (When 350 oC is not

reached welding is performed continuously by turning

In addition, a comparison of the interpass temperature

histories is indicated in Fig 4 The interpass temperature

histories shown here are recorded at the measuring point

recommended by JASS 6; in more detail, the temperature

histories at the nodal point at a position on the base

metal side 10 mm from the groove upper end zone, which

is the centre of the weld length It is apparent from this

figure that the interpass temperature prior to the start of

the welding of the final pass increased to approximately

500 oC in the continuous uni-directional and the

continuous turning back welding types By contrast, it

is clear that the interpass temperature of the controlled

welding type did not exceed 350 oC due to the introduction

of the weld dwell prior to welding the two final layers

The history of the interpass temperature and the weld

heat input are as indicated above and it was demonstrated

that the idealised state of the weld heat input conditions

during actual welding procedure can be reproduced by

the analytical conditions established by this study

Effects of welding conditions upon

microstructure and hardness distribution

Effects of deposition sequence upon

microstructural distribution

Figure 5 shows the bainitic phase proportion near the

weld metal following the completion of welding using the continuous turning back and the interpass temperature controlled welding types; the figure indicates examples

of the mircrostructural distributions, obtained following the analysis using each deposition sequence The cross-sections at the centre of the weld length are indicated According to the figure the bainitic phase proportion of the weld metal zone was approximately 70–80 % for the continuous turning back welding type and more than

90 % for the controlled welding type Furthermore, it is apparent that the continuous turning back type weld has a greater HAZ width It is surmised that these discrepancies are caused by the variations of temperature history and interpass temperature due to the difference

in the deposition sequence That is to say, the heat input conditions and the cooling characteristics of the weld zone vary according to the difference of the interpass temperature; consequently, it is surmised that differences will arise in the microstructures in the weld zone and the HAZ and even in the hardness, in other words, the strength Consequently, in this section, the weld metal zone and the HAZ were considered and in particular, an investigation was made into the variation in the heat input conditions and the cooling characteristics of the weld zone due to the difference in the interpass temperature

Effects of deposition sequence upon the cooling characteristics and hardness of the weld metal

In the first instance, attention was given to the weld metal centre zone and differences in the Vickers hardness were identified and compared for the three weld types of continuous unidirectional welding, continuous turning back welding and interpass temperature controlled welding

in which the deposition sequence of the weld pass was varied between the first layer and the last layer Figure 6 shows the results The centre zone of the boundary between the first and the second layers was described

as the first layer side and the centre zone of the boundary between the sixth and the seventh layers was described

as the last layer side; the Vickers hardness at points aF,

aL and cF, cL 10 mm from the start and the finish zones of the weld, and at points bF, bL at the centre zone in the axis of the weld line are indicated for both

It is evident that there is a difference in the strength corresponding to the Vickers hardness due to the difference in the welding conditions but there is little

4 Comparison of temperature histories during weld process

at recommended points by JASS 6.

5 Distributions of bainitic phase proportion near weld metal.

Bainite 0 0.1 0.2 0.3 0.4 0.5 0.6 0.8 0.0 1

difference in the strength along the weld line when

identical welding conditions are employed The maximum

temperature at the weld metal zone is not affected by the

welding conditions so the microstructure and the strength

of the weld metal are determined by the cooling rate

Consequently, for this study, a cooling time t8/5 from

800 oC to 500 oC, which is extensively employed for

common structural steel materials, was used as an indicator

to evaluate the mechanical properties

First, the Vickers hardness at the last layer side was

investigated At points aL, bL, cL to which attention was

paid on the last layer side, the microstructure turns into

austenite due to the last pass welding; this implies that

it is acceptable to consider the cooling process of the

last pass only Accordingly, the cooling process of the

last pass was observed; Fig 7 shows the results of the

investigation into the distribution of the cooling time

t8/5 in the weld metal along the weld line The figure

indicates that the interpass temperature controlled

welding type has a significantly short cooling time t8/5

compared with the continuous uni-directional welding

type and the continuous turning back welding type, in

other words, it has a faster cooling rate This is assumed

to be the cause of the variation in the strength at the last

layer side of the weld metal zone The increased cooling

rate, where the interpass temperature is controlled, is

caused by the effect of the increased temperature gradient

since the preheating is similar to that when welding at

low temperature In addition, the continuous unidirectional

welding type has a greater cooling time t8/5 than the

continuous turning back welding type; it is surmised

that this is due to the tendency for even heat input over

the whole joint during continuous unidirectional welding compared with continuous turning back welding which provides heat more intensively For this analysis of the continuous uni-directional welding it was assumed that there was no welding torch travel time from the finish end to the start zone; however, during the actual welding process a certain time is required for the suspension and the restart of the welding arc, so cooling continues for the entire weld during this period; it is surmised that the consequences are that there is an increased possibility that the cooling time t8/5 might be less than for the turning back welding

Furthermore, from observation of the distribution along the weld line it can be seen that t8/5 is large at the positions where there is a turning back operation, for example, at the left end of the continuous turning back welding type and the right end of the continuous multi-pass welding type; in other words, the cooling rate is reduced at these positions In addition, a tendency can be seen such that the cooling rate becomes slightly increased at the weld start and finish zones It has been reported that the cooling rate of the start and the finish of an infinite sheet weld becomes approximately two times the cooling rate of the centre zone along the weld line under quasi-steady state conditions.18 This study employs multi-layer finite thick plate welding, so the cooling rate is not as fast as that for infinite sheet metal However, variations in t8/5 are observed along the weld line but these are not sufficiently large to cause major differences in the Vickers hardness, that is to say the strength, and it is assumed that the difference in the cooling rate along the weld line hardly affects the strength of the last layer side of the weld metal

Next, the significance of the cooling rate to the Vickers hardness of the last layer side of the weld metal was studied The vertical axis in Fig 8 shows the plotted Vickers hardness of every point on the last layer side of the weld metal zone, as indicated in Fig 7, and the horizontal axis indicates the plotted cooling time t8/5 from

800 oC to 500 oC The equation for calculation of the Vickers hardness of each microstructure employs the chemical

6 Comparison of Vickers hardness in weld metal.

7 Distribution of elapsed cooling time t 8/5 in weld metal at last layer side.

composition and the cooling rate; thus it becomes evident

that the Vickers hardness increases linearly with increasing

cooling rate; it is possible to deduce this by consideration

of the mechanisms and it is feasible to evaluate

quantitatively the Vickers hardness, that is the strength,

of the last layer side of the weld metal zone using t8/5

That is to say, the reduction of the heat input and the

subsequent restraint of the interpass temperature to a

low value for the last layer side of the weld metal zone

are considered to be effective to ensure the strength by

means of employing interpass temperature controlled

welding and also using l layer multi-pass welding

Next, the Vickers hardness of the first layer side of

the weld metal zone was considered; the weld metal of

the first layer side was subjected to the multiple thermal

cycles due to the heat inputs of the succeeding weld

passes and the continuous cooling transformation diagram

cannot be applied simply as it can be for the last layer

side However, the general tendency of the Vickers hardness distribution of the first layer side is noted to

be approximately similar to that of the last layer side That is to say, comparison of continuous uni-directional welding, continuous turning back welding and interpass temperature controlled welding at the respective positions showed the Vickers hardness of the interpass temperature controlled type welding became high in every case; such

a tendency was noted to be similar to that of the last layer side Furthermore, the tendency was seen to be generally similar for the start and finish zones and the centre zone and this was also similar to that of the last layer side It is assumed from the above that, even for the first layer side of the weld metal zone, restraint of the interpass temperature to a low value is effective in order to increase the cooling rate and subsequently to ensure the strength, similarly to the case of the last layer side

It becomes necessary to obtain the cyclic heating cooling transformation diagram for the multiple thermal cycles, consisting of various patterns, in order to apply the continuous cooling transformation diagram to the regions which are subjected to complex multiple thermal cycles such as the first layer side weld metal It is essential

to make an attempt to evaluate this in detail by systematically gathering the data of such complex multiple thermal cycles; in addition, it is thought that the establishment of some sort of simplified evaluation technique would become an industrially valuable tool

Effects of deposition sequence upon cooling characteristics and hardness of the HAZ

The width of the distribution of the HAZ is assumed to

8 Relation between Vickers hardness and elapsed cooling

time t 8/5 in weld metal at last layer side.

9Comparison of Vickers hardness near heat-affected zone

be affected by the deposition sequence, so an

investigation was carried out into the Vickers hardness

to a distance of 10 mm from the weld bond zone on the

diaphragm side of the first layer side and the last layer

side This was determined after taking into account the

regions to which the greatest moments act when an

earthquake load is applied Figure 9 was produced,

showing the groove vertical plane as the origin, with the

aim of investigating Vickers hardness variations; the

boundary between the first layer and the second layer

was called the first layer side and the boundary between

the sixth layer and the seventh layer was called the last

layer side The figure shows the Vickers hardness at the

cross sections A and C, 10 mm from the weld end zone,

and the cross section B at the centre of the weld length

for the first layer side and the last layer side The following

discussion assumes that the range over which hardening

occurred became the HAZ width

Firstly, continuous unidirectional welding and

continuous turning back welding were compared The

Vickers hardness in the vicinity of the groove vertical

plane varied greatly along the weld line Observation of

the cooling time t8/5 from 800 oC to 500 oC at the HAZ

indicated that the cooling rate at the weld start zone of

the continuous uni-directional welding type was fast and

the cooling rate at the start and the finish weld zones of

the continuous turning back welding type was slow; it

is evident that this caused the difference in the Vickers

hardness along the weld line

Furthermore, it is apparent that the HAZ widths at

any cross-section of the last layer side of the continuous

uni-directional weld and the continuous turning back

weld were approximately similar but the HAZ widths at

the end cross sections A and C on the first layer side

increased for the continuous turning back welding type

Approximately similar heat input patterns were noted and

there was little difference in the HAZ width at the centre

zone along the weld line for both the continuous

uni-directional welding type and the continuous turning back

welding type, so this is surmised to be due to the effects

of the turning back operation at the weld start and finish

zones Figure 10 shows the temperature distribution during the turning back operation at the cross section C of the first layer side where a turning back operation was carried out during the transition from the first layer to the second layer The broken line indicates the temperature distribution at the time when the heat source arrived during the first layer welding and an identical temperature distribution was noted for both types Subsequently, during second layer welding, the heat source for the continuous uni-directional welding type moved to the cross section A side, so cooling took place; by contrast,

in the continuous turning back welding type reheating took place due to the turning back operation, so the temperature increased further; it is assumed that the HAZ width was increased as a result of this operation The effects of the turning back operation cause differences in the temperature distribution and the cooling characteristics at the weld start and finish zones; as a result, the HAZ width varies; however, there was no great difference evident in the heat input and the interpass temperature history for both, as indicated in Figs 3 and

4 This implies that, for welding with the control of the interpass temperature at the centre of the weld length and with the heat input specified in JASS 6, there are some cases which do not take into account the effects

of the turning back operation at the weld start and finish zones

Next, cases of the continuous turning back welding and the interpass temperature controlled welding were compared First, the differences in the Vickers hardness near the weld bond zone were examined The cooling characteristics at the last layer side were noted as similar

to those of the weld metal zone, so the difference in the Vickers hardness corresponded to the weld metal zone

By contrast the Vickers hardness declined at the weld start and finish zones of the first layer side compared with the centre of the weld length due to the effects of the turning back operation as described in the previous section Furthermore, the differences in the Vickers hardness (near the bond zone of the weld start and finish end zones of the first layer side) between the continuous turning back welding type and the interpass temperature controlled welding type are caused by the varied cooling characteristics depending upon the action of interpass temperature control; this is similar to the case of the weld metal zone

In addition, as evident from the fact that there was no difference in the heat input pattern and the interpass temperature history for the first layer side, the HAZ width was seen to be approximately similar However, it was clear that the HAZ width in the last layer side was reduced for the interpass temperature controlled welding type Figure 11 shows the distribution of the maximum temperature at cross section B which is at the centre of the weld length on the last layer side The figure indicates the positions where the maximum temperature reached

800 oC and it was approximately 6 mm from the groove face for the continuous turning back welding type and approximately 4 mm from the groove face for the interpass temperature controlled type; it is evident that these values

10 Effect of turning back operation on temperature

distribution near heat-affected zone at first layer side of

turning back side.

correspond to the HAZ width finally obtained.

As described, comparison of the state of the HAZ for

continuous turning back welding and for interpass

temperature controlled welding shows that there is no

great difference in the HAZ width at the first layer side

and that the interpass temperature control has no effect

It is assumed that this is because there is no requirement

for a weld dwell at the first layer side for the interpass

temperature control which has an upper temperature limit

of 350 oC at the measuring point recommended by JASS

6 On the other hand, the HAZ width of the last layer

side decreases when the interpass temperature is

controlled However, when controlling the interpass

temperature, attention should be given to the increased

weld heat input near the last layer side and a slightly

increased hardening at the HAZ Furthermore, as described

in the previous section, the effects of interpass temperature

control by the JASS 6 system were seen in the weld metal

zone, irrespective of the first layer side or the last layer

side and the effects of the interpass temperature control

upon the weld metal zone and the HAZ were varied

It was seen from the results obtained in this report

that the welding procedure control as stipulated in JASS

6, especially the interpass temperature control, is effective

to a certain extent It is obviously important, when

performing the welding procedures, to understand fully

how the heat input and the interpass temperature change

with the procedure control and, as a consequence, how

they affect the metallurgical/mechanical properties near

the weld zone On the other hand, there are some aspects

of joint performance which are probably not sufficiently

determined only by the control using existing JASS 6

and it is thought that, in the future, it will be necessary

to develop a simplified control method which also takes

into account the welding procedure efficiency

Finally, the hardness distribution at each zone of the

welded joint could be obtained by carrying out the

temperature/microstructure simulation of the welding

process, as described in this report This implies that it

is feasible to carry out the tensile strength prediction

for each region of the welded joint using the analytical

results of the temperature and the microstructure.19,20 In

the future, it is thought that a practical approach can be obtained by refining the hardness equation and the accuracy of the input material properties and by further improvement of the analytical accuracy

Conclusions

The temperature/microstructure/hardness simulation was carried out for the welding of beam-to-column welded connections of steel framed strcutures; an investigation was performed on the joint microstructure and hardness changes due to the difference in the temperature history The temperature history, the microstructure and the hardness of various deposition sequences were obtained analytically using a three-dimensional multi-layer welding simulation, which takes the moving heat source into consideration; it was demonstrated that the relationship between the cooling rate and the hardness of the weld zone can be explored by the use of the cooling time t8/5 from 800 oC to 500 oC as a parameter indicating the cooling rate Also, a further investigation was carried out on the effect of the deposition sequence upon the HAZ width and the Vickers hardness In addition, the feasibility of prediction of the tensile strength of the welded joint which had been subjected to the thermal cycles was demonstrated using the results of the temperature/ microstructure analysis

References

11 Effect of interpass temperature control on maximum

temperature near heat-affected zone at last layer side of

centre of weld length.

1 Bertero VV et al: ‘Performance of steel building structures

during the Northridge Earthquake’ The Earthquake Research Center, University of California at Berkeley, Report No UCB/ EERC-94-09 (1994).

2 Toyoda M: ‘How steel structures fared in Japan’s great

earth-quake’ Welding Journal 1995 74 (12) 31–42.

3 Mochizuki M et al: ‘Residual stress reduction and fatigue

strength improvement by controlling welding pass sequences’.

Transactions of the American Society of Mechanical Engi-neers, Journal of Engineering Materials and Technology 2000

122 (1) 108–112.

4 Mochizuki M et al: ‘Residual stress distribution depending on

welding sequence in multi-pass welded joints with X-shaped groove’ Transactions of the American Society of

Mechani-cal Engineers, Journal of Pressure Vessel Technology 2000

122 (1) 27–32.

5 Mukai A et al: ‘Investigation on MAG wires for building

struc-ture’ Journal of Steel Construction Engineering 2000 7 (26)

1 3 – 2 5

6 Nakagomi T: ‘Control of interpass temperature in

beam-to-column structures of steel frame structures’ Structural

Tech-nology 2001 14 (5) 18–25.

7 Furuya H et al: ‘Newly proposed HAZ toughness evaluation

method of structural steel for building and the result of

evalu-ation by its applicevalu-ation’ Journal of Steel Construction

Engi-neering 2001 8 (32) 17–31.

8 Okazawa T et al: ‘Study on influence of heat input and

inter-pass temperature on the mechanical properties and the

execu-tion efficiency of weld joints in steel structures’ Journal of

Steel Construction Engineering 2001 31 (3) 1–10.

9 Mochizuki M and Toyoda M: ‘Effects of welding multiple heat

cycles on joint performance in beam-to-column welded con-nection of steel framed structures – studies on relation be-tween joint performance and welding heat cycles in welded

joints (Report 1)’ Quarterly Journal of Japan Welding

Soci-ety 2003 21 (1) 46–53.

1 0 Architectural Institute of Japan, edited: Japanese

Architec-tural Standard Specification (JASS 6) – Steel Work, 7th Edi-tion’ Architectural Institute of Japan, Tokyo, Japan 1996.

11 L e b l o n d J B a n d D e v a u x J : ‘ A n e w k i n e t i c m o d e l f o r

anisothermal metallurgical transformation in steels including

effect of austenite grain size’ Acta Metallurgica 1984 32 (1)

1 3 7 – 1 4 6

1 2 Blondeau R et al: ‘Forecasting the hardness and resistance of

carbon and low-alloy steels according to their structure and

c o m p o s i t i o n ’ L e s M é o i r e s S c i e n t i f i q u e s d e l a R e v u e d e

Métallurgie 1973 70 (12) 883–892.

1 3 Blondeau R et al: ‘Forecasting the hardness, resistance and

breaking point of carbon and low-alloy steels according to

t h e i r c o m p o s i t i o n a n d h e a t t r e a t m e n t ’ L e s M é o i r e s

Scientifiques de la Revue de Méallurgie 1975 72 (11) 759–

7 6 9

1 4 SYSTUS International: SYSWELD 2000 Reference Manual

SYSTUS International 2000.

1 5 Mochizuki M et al: ‘Welding residual stresses at the

intersec-tion of a small diameter pipe penetrating a thick plate’

Nu-clear Engineering and Design 1993 144 (3) 439–447.

1 6 Mochizuki M et al: ‘Numerical analysis of welding residual

stress and its verification using neutron diffraction measure-ment’ Transactions of the American Society of Mechanical

Engineers, Journal of Engineering Materials and Technology

2000 122 (1) 98–103.

1 7 Mochizuki M et al: ‘Generation of thermal and residual stresses

during welding heat cycles – Studies on residual stress in welded joints considering phase transformation effect (Report 1)’.

Quarterly Journal of Japan Welding Society (to be published).

1 8 Satoh K et al: Welding Engineering Rikogaku-sha 1979 39–

4 2

1 9 Voss O et al: ‘Consideration of microstructural

transforma-tion in the calculatransforma-tion of residual stresses and distortransforma-tion of

larger weldments’ Mathematical Modelling of Weld

Phenom-ena 4, IOM Communications, London 1998 584–596.

2 0 Karkhin V A et al: ‘Effect of low-temperature phase

trans-formation on residual stress distributions in laser welded joints’.

Mathematical Modelling of Weld Phenomena 5, IOM

Com-munications, London 2001 597–614.