The Leveling Property of the Weld Pool

I. G. Lyapichev

(Presented by Prof. A. N. Dobrovidov, Doctor of Technical Sciences)

A constant chemical composition of the deposited metal is the basic condition for obtaining invariant physical and mechanical properties of the weld. This is necessary both in welding structural steels and, especially, in surfacing highly alloyed steels.

To obtain a constant chemical composition of the deposited metal, it is necessary to maintain an unchanging ratio of the amounts of molten flux, ferroalloys, base metal, and electrode metal per unit time. Under plant conditions this requirement is often violated because of fluctuations in the line voltage and the nonuniformity of feeding and granulation of the flux and ferroalloys.

The conditions for obtaining a homogeneous chemical composition of the deposited metal have not been sufficiently studied. It has not been clarified, for example, what effect the mixing of the metal in the weld pool has on the homogeneity of the metal composition.

The first confirmation of the mixing of metal in the weld pool (in manual welding) was discovered by A. A. Alov [1]. Silver introduced into the lower part of the weld pool through the base metal was found at a considerable distance from the point of introduction. In addition, it was located in the upper layer of the deposited metal, despite the fact that the specific gravity of silver is greater than the specific gravity of iron.

The presence of intensive mixing of metal in the weld pool has repeatedly been noted by G. I. Pogodin-Alekseev [2], K. K. Khrenov [3], and others.

A constant chemical composition of the bead can be ensured only with an unchanged linear power of the arc.

Any change in the linear power of the arc causes a disturbance of the established heat balance, which causes fluctuations in the composition of the bead. When low-carbon welding wire is used for welding low-carbon steel, a change in the linear power of the arc is less dangerous, since the base and electrode metals have approximately the same chemical composition.

In surfacing high-speed steel we used low-carbon welding wire. Alloying of the deposited metal was carried out with ferroalloys.

Ferroalloys may be uniformly distributed in the flux¹ (alloying fused flux) or concentrated directly in the arc-burning zone in the form of an alloying paste [9] or a bar made of ferroalloys. However, the use of paste is associated with a number of technological difficulties: filling the groove for surfacing with paste, the need for subsequent drying of the workpiece with the applied paste. Storage of the paste is difficult—

(since it “sets” and hardens), its use in surfacing flat workpieces (tool holders for cutters, knives for inserted-tooth milling cutters, etc.), when edge preparation for surfacing may be absent. Dosing the paste according to the profile of the preparation cannot provide the required accuracy and, consequently, constancy of chemical composition.

The alloying of the surfaced metal with bars of ferroalloys, as proposed by us, is free of these shortcomings. The influence of fluctuations in the line voltage on the chemical composition of the surfaced metal is then considerably weaker than with other methods of alloying.

Alloying unmelted fluxes consist of slag-forming and alloying components. When deoxidation of the surfaced metal is insufficient, or when there is a large loss of alloying elements by oxidation, deoxidizers are introduced into the flux composition. A method for calculating an alloying unmelted flux or the composition of a ferroalloy bar has not yet been developed. The amount of ferroalloys that must be introduced into the slag-forming part of the flux or concentrated in the form of a bar is determined by a series of trial surfacings. The composition found for the alloying flux or ferroalloy bar ensures the required composition of the bead while maintaining the trial surfacing conditions. A change in even one parameter of the surfacing conditions \((J_{\mathrm{d}}, U_{\mathrm{d}}, V_{\mathrm{sv}}, V_{\mathrm{e}}, d_{\mathrm{e}}, U_{\mathrm{seti}})\) causes a disturbance in the ratio of the amounts of melted flux, base metal, and electrode metal and, consequently, a change in the chemical composition of the surfaced metal.

The ferroalloys introduced must alloy the melted base and electrode metal.

It is assumed here that the base metal is carbon or low-alloy steel, and the electrode is low-carbon wire of grade Sv-08. As a first approximation, independently of the composition of the slag-forming base of the welding flux, the amount of ferroalloys introduced may be calculated from the linear weight of the bead of surfaced metal. In this case, the magnitude of ferroalloy burn-off remains unaccounted for, as does the portion of them left unused in the solidified slag crust, if the alloying of the surfaced metal is carried out with a ceramic flux.

Fluctuations in the linear power of the arc when surfacing high-speed steel have a stronger effect on the chemical composition than when welding low-carbon steel with low-carbon wire. This occurs because the change in the amount of flux, ferroalloys, base metal, and electrode metal melted separately is not proportional to the change in arc power. The degree of this nonproportionality depends on the type of welding equipment and on the method of alloying the surfaced metal.

When alloying the surfaced metal through a ceramic flux, the chemical composition of the bead depends to a greater extent on voltage fluctuations than with other methods of alloying.

Concentrating the ferroalloys in the form of a bar in the arc-burning zone ensures their complete melting independently of fluctuations in the linear power of the arc. A change in the amount of melted protective flux has little effect on the composition of the bead. With this method of introducing ferroalloys, in order to obtain a uniform chemical composition that depends little on fluctuations in the linear power of the arc and on the type of welding equipment, it is necessary to ensure a constant amount of melted base and electrode metal.

Fluctuations in the linear power of the welding arc are caused for the most part by changes in the line voltage.

Experimental verification of the influence of line-voltage fluctuations within the range of \(200 \div 220\ \mathrm{V}\), of the alloying method, and of the type of welding equipment on the chemical composition of the surfaced metal showed that the great-

…greater constancy of the chemical composition is provided by welding equipment with a constant electrode feed rate (the TS-17 tractor) when surfacing the deposited metal with bars made of ferroalloys. This is explained by the fact that, when the mains voltage fluctuates, the TS-17 tractor ensures a constant amount of melted electrode metal. When the mains voltage fluctuates, a new regime is established with a reduced arc voltage and a practically unchanged welding current. Despite the overall decrease in arc power when the mains voltage drops, the amount of melted base metal changes insignificantly. This occurs because the arc column is immersed in the base metal, which increases the share of heat expended on melting the base metal.

Our experiments agree with the theoretical investigations of welding equipment by I. L. Brinberg [7]. The experiments performed show the possibility of using welding heads with a constant electrode feed rate for surfacing high-speed steel when the mains voltage fluctuates: \(210 \pm 10\) V. If the head is adjusted at the highest mains voltage and only its decrease is expected, it is necessary to calculate the ferroalloy bar according to the lower limit of the content of elements in the steel. A drop in voltage will reduce the amount of melted base metal and will cause an increase in the alloying content of the bead.

It should be noted that the permissible fluctuations according to GOST in the content of elements in steel R18 reach 8.5–25%, and in steel R9—11.8–30%.

Heads with a regulated electrode feed rate (ADS-1000), when the mains voltage drops, establish a new regime with unchanged arc voltage and reduced welding current. The latter leads to a decrease in the melted electrode and base metal. The amount of melted flux decreases significantly. The composition of the deposited metal changes more strongly than when using heads with an independent, constant electrode feed rate. This is explained by the unfavorable response of the ADS-1000 head to fluctuations in mains voltage, which leads to a change in a technologically important regime parameter (welding current) and to preservation of the secondary (arc) voltage.

Equipping heads with a constant electrode feed rate with primary-voltage stabilizers will increase their value in surfacing operations.

As the experiments showed, the greatest stability of the composition of the deposited metal when the mains voltage fluctuates is ensured by using automatic machines with an independent electrode feed rate and by surfacing with bars made of ferroalloys. Regardless of fluctuations in the mains voltage, these machines ensure a constant amount of melted electrode and base metal and ferroalloys. The latter is facilitated by the fact that the ferroalloys are not distributed throughout the slag-forming part of the flux, but are concentrated directly in the surfacing zone.

A change in the amount of melted slag-forming part of the flux has little effect on the composition of the deposited metal, since alloying is carried out by the ferroalloy bars. At the same time, it remains unclear how the nonuniformity of ferroalloy dosing in the bar affects the chemical composition of the deposited metal along the length of the bead.

Method for calculating the composition of a ferroalloy bar

The composition of the deposited metal depends on the amount of melted ferroalloys in the bar, base metal, electrode metal, and flux. The calculation of the composition of the ferroalloy bar must be carried out depending on

surfacing conditions and flux composition. Fluctuation of the surfacing conditions leads to a change in the amount of molten base and electrode metal, ferroalloys, and flux and, consequently, to a change in the composition of the bead.

These requirements are not fully satisfied by the method for calculating the composition of the alloying paste proposed by I. G. Kosmachev [9]. In this method, a large reserve (more than 30%) of alloying elements is provided without relation to their burn-off. The dosage of the paste is made as a function of the cross-sectional area of the grooves for surfacing. There is no connection with the surfacing conditions and the flux grade. Use of the calculation method is difficult when complex or carbonaceous ferroalloys are used.

The method we propose for calculating the composition of a bar made of ferroalloys is based on the assumption that the cross-sectional area of the bead, at constant linear arc power, does not depend on its chemical composition. The cross-sectional area of the bead when surfacing over bars made of ferroalloys and without them will be the same. When surfacing over bars, the amount of molten base metal of low-carbon steel must decrease in proportion to the weight of the ferroalloy bar.

A change in the chemical composition of the deposited metal and, consequently, in the heat of fusion will lead only to a change in the degree of superheating of the metal of the weld pool. However, this does not lead to a corresponding fluctuation in the cross-sectional area of the bead because of insufficient heat transfer from the liquid to the solid metal.

To determine the influence of the density and chemical composition of the bar on the cross-sectional area of the bead, the following experiments were carried out. From chips of steels 3 and R18 and from mixtures of ferroalloys of various composition, bars of different density but of the same weight—26 g—were made by pressing, with dimensions \((6 \div 8)\times 16\times 50\) mm; from steel 3, plates \(4\times 16\times 50\) mm were made, and from steel R18, plates \(3.6\times 16\times 50\) mm, each weighing 26 g. Surfacing was carried out with a TS-17 tractor on blanks \(18\times 8\times 370\) mm of steel 3 under the following conditions: \(J_d = 600\) A, \(U_d = 45\) V, \(d_e = 4\) mm. \(U_{\text{mains}} = 208\text{–}210\) V, \(U_e = 40\) m/hour, \(V_{\text{sv}} = 20\) m/hour under AN-348-A flux, using pressed bars, plates, and without them, i.e., directly on the blanks.

Measurements of the cross-sectional areas of the beads showed the absence of a dependence on the chemical composition and density of the alloying addition (bar or plate). The cross section of the bead when surfacing over a plate and chips of steel 3 was equal to \(1.19\text{–}1.23\ \text{cm}^2\), over a plate and chips of R18—\(1.14\ \text{cm}^2\), and over a mixture of ferroalloys—\(1.09\ \text{cm}^2\). If the different specific gravity of the deposited metal is taken into account, it turns out that the weight of the molten base and electrode metal is constant and does not depend on the composition of the bead. In conclusion, we carried out the following experiments.

In the TS-17 tractor, adjusted to the above conditions, the low-carbon welding wire Sv-08 was replaced by silver steel \(\varnothing 4\) mm made of steels R9 and R18. The base metal in this case was also blanks of steels R9 and R18. No disturbance of the stability of the surfacing process was observed. The cross-sectional area of the deposited metal was \(1.18 \div 1.24\ \text{cm}^2\).

We recommend carrying out the calculation of the composition of a bar made of ferroalloys in the following sequence:

1. The welding flux is selected beforehand. As experience has shown, high-manganese fluxes of the AN-348-A type may be used. The physicochemical features of alloying metal deposited with the use of ferroalloy bars ensure high quality of the deposited high-speed steel and insignificant burn-off of the alloying elements.

2. According to the drawing of the part to be surfaced, or of the cutting part of the tool, a surfacing regime is specified that provides the necessary transverse—

new cross section of the deposited metal. The surfacing conditions may be selected according to the data of work [10].

1. Trial surfacing is carried out on a workpiece of the part or tool with low-carbon welding wire under the selected conditions.

2. The cross-sectional area is measured and the linear weight of the bead is determined.

3. If the bead cross section satisfies item 2, then, by formula (1), the composition of the ferroalloy bar is calculated for alloying the deposited metal to the required level.

The amount of ferroalloy \(Q_2\), the introduction of which into \(A_2\) of deposited metal will ensure in it a content of \(a\%\) of the element, we recommend determining by the formula:

\[ Q=\frac{100+k}{100}\,A\,\frac{a}{b}, \tag{1} \]

where

\(b\) — content of the element in the ferroalloy, in %,
\(k\) — coefficient taking into account burn-off and other losses, in %.

If \(c\) of the element is introduced, in addition to the ferroalloy, for example by slag-forming components, electrode metal, or base metal, then in the formula, instead of “\(a\)” one should substitute “\(a-c\).”

1. The ferroalloy bar is made in a mold. The dimensions of the bar are established as follows: the length of the bar must provide the linear dosing of ferroalloys. The width of the bar is taken as 2 mm less than the width of the bead of deposited metal (item 3). After manufacture, the bar is placed on the surface of the part being surfaced or of the cutting-tool blank. After this, surfacing is carried out under the previously selected conditions (item 1). The molten bar alloys the weld pool.

Under the surfacing conditions:

1) line voltage \(U_{\text{line}} = 210\ \text{V}\),

2) welding current — \(J_0 = 600\ \text{A}\),

3) arc voltage — \(U_v = 45\ \text{V}\),

4) surfacing speed — \(V_{\text{sv}} = 20\ \text{m/hour}\),

5) electrode diameter — \(d_e = 4\ \text{mm}\),

6) welding equipment TS-17, flux AN-348-A,

to make a bar measuring \(5 \times 16 \times 50\ \text{mm}\), 26 g of mixture was used: 14.3 g of ferrotungsten (70% W, 0.8% C); 9.0 g of ferrochrome (52.2% Cr, 6.02% C, 3.21% Si); 6.75 g of ferrovanadium (39.8% V, 0.52% C, 2.8% Si, 2.50% Al); 3.1 g of ferromolybdenum (59.5% Mo, 1.42% Si), and 19 g of cast iron (4.81% C). The cast iron was melted on a base of U12 steel.

Melting of U12 steel scrap was carried out on an AZ-43 high-frequency generator at a voltage of 8–9 kilovolts and 2.5–3 amperes of anode current with a reduced frequency due to the additional introduction of a bank of air capacitors. A crucible made from a mixture of 80% quartz sand and 20% refractory clay was placed in the inductor of the unit. The capacity of the crucible was 1.5–2 kg of steel.

Carburization of the steel was carried out with graphite powder. The melt was protected with powdered window glass.

Alloying with ferroalloy bars of this composition ensured the production of deposited metal whose chemical composition is given in Table 1.

To prevent lack of fusion, the linear weight of the ferroalloy bar must not exceed 80% of the linear weight of the bead obtained by surfacing with low-carbon wire (item 4).

Alloying of the deposited metal with ferroalloy bars opens up a broad possibility for using briquettes made from chips of alloy steels. In this case a considerable saving of ferroalloys is achieved.

The use of briquettes made from chips of special steels for alloying the deposited metal is based on the fact that, for each automatic surfacing condition, there is a maximum amount of ferroalloys (in grams) that can be introduced per unit length of deposited metal. If this amount is exceeded, lack of fusion is observed. For example, this occurred at a ferroalloy dosage of 9.2 g/cm when surfacing under the above-mentioned conditions: welding current—600 A, electrode-wire diameter 4 mm.

Earlier experiments on surfacing high-speed steel using ferroalloy bars \((I_0 = 600\text{ A},\ V_{\text{w}} = 20\ \text{m/h},\ V_{\text{e}} = 40\ \text{m/h},\ d_{\text{e}} = 4\ \text{mm})\) show that, in the bead composition, the proportion of electrode metal does not exceed 20%. The proportion of electrode metal was determined from the cross-sectional area and the linear weight of the bead, as well as from the consumption of electrode wire.

By increasing the amount of ferroalloys introduced, it is possible to reduce the depth of penetration of the base metal to 1–1.5 mm. This will reduce dilution of the weld pool by the low-carbon steel of the part or tool blank. The proportion of base metal in the weld pool will decrease to 15–20%. As a result, 35–40% of the deposited metal will consist of unalloyed base and electrode metal. This makes it possible, for example, to obtain deposited metal whose chemical composition is close to steel of the R9 type, if a briquette of chips from R18 steel is used. With a linear weight of the bar made from R18 steel chips of 6.5–7 g/cm, when surfacing under the indicated conditions with AN-348-A flux (TS-17 tractor), we obtained the following composition of the deposited metal: 0.46% C; 0.84% Mn; 0.70% Si; 10.02% W; 1.82% Cr and 0.54% V. The reduced content of chromium, vanadium, and carbon can be compensated by adding ferroalloys and cast iron. The results of such combined alloying with chips and ferroalloys are given in Table 1.

Table 1

When alloying the deposited metal with chips and ferroalloys, we used 40 g of R18 steel chips; 4.4 g of ferrochrome; 4.8 g of ferrovanadium; and 26 g of cast iron based on R18 steel waste. Such replacement of the ferroalloy bar, consisting of 14.3 g of ferrotungsten, 9.0 g of ferrochrome, and 6.75 g of ferrovanadium, saves 100% ferrotungsten, 51% ferrochrome, and 30% ferrovanadium. The use of additions of chips from high-chromium steel of the Kh12 type will lead to an increase in the saving of ferrochrome.

Further work should expand the field of application of chips from alloy steels, both in the form of briquettes and for the manufacture of alloy-

... of alloying ceramic fluxes. The use of special surfacing modes that ensure the least penetration of the base metal, as well as equipment with a nonconsumable (graphite, tungsten) electrode, will significantly increase the concentration of alloying elements in the weld pool.

At present, at the Tomsk Order of the Red Banner of Labor Cutting-Tool Plant, tool surfacing is being introduced with the use both of ferroalloys and of chips of high-speed steel.

The chips are briquetted on a hammer or friction press (180 t). The chips are first moistened with liquid glass. A measured amount of 70 g of chips is poured into the die. After pressing, the briquettes are calcined at a temperature of 350–400° for 1 hour. The weight variation of the finished briquettes reaches 64–72 g. The length of the briquettes is 100 mm, the cross-section 6 × 16 mm.

Alloying with bars made of ferroalloys was used by us to obtain experimental surfacings of high-speed steel of the R9 type, additionally alloyed with carbon, vanadium, titanium, and tungsten.

To determine the required fineness of the ferroalloys (granulation), the bars were made from a mixture of their powders passing through a sieve with an aperture diameter of 0.5–2 mm.

After mixing the ferroalloys in the dry state, the mixture was slightly moistened with liquid glass previously diluted with 50–60% water. After this, repeated mixing is carried out.

A slight moistening of the mixture with diluted liquid glass slows the “setting” and hardening of the mixture. In addition, the channels between the particles of the mixture remain free for the escape of vapors during drying. The latter makes it possible to avoid swelling of the bar during accelerated drying and to shorten it to 2–3 hours by increasing the temperature to 80–100°. The bonding strength of the mixture particles is quite sufficient: the bars are not destroyed during transportation and light impacts.

The moistened mixture is poured into the press mold and leveled. Dosing of the mixture during filling may be performed either by a measuring container or by weighing. Pressing is carried out at a low pressure—8–10 kg/cm². On external inspection and destruction of the bars, no voids were found.

Fineness of the ferroalloys up to 2 mm does not affect the technological properties of the flux or the uniformity of the chemical composition of the deposited metal. Such an increase in the coarseness of the ferroalloy powder, instead of grinding it and screening it through a sieve with 1600 holes/cm² [9], increases the productivity of the crushing equipment.

The principal advantage of surfacing over bars made of ferroalloys, compared with surfacing under alloying ceramic fluxes, consists in the possibility of obtaining a highly alloyed composition (up to 30–40%) of the deposited metal in the first bead, i.e., in single-layer surfacing. In this case it is possible to use any modes of automatic welding that ensure the greatest cross-sectional area of the deposited metal. Significant melting of the base metal when using modes with high welding-current densities is not an obstacle to the introduction of this method of surfacing. On the contrary, this leads to the possibility of increasing the alloying of the surfacing, since the linear weight of the bar made of ferroalloys is in direct dependence on the amount of molten base metal.

When alloying the deposited metal with a ceramic flux, its calculated composition can be obtained only in the second or third layer of metal [4, 6].

The transfer of alloying elements from the ceramic flux into the deposited...

the deposited metal does not exceed 60% because of considerable losses of ferroalloys in the slag crust and oxidation.

If one takes into account that the technology for making bars from ferroalloys and from ceramic flux is approximately the same, then, in single-pass surfacing of beads of high-alloy steel with a large cross-sectional area, alloying by means of bars has obvious advantages.

In surfacing high-speed steel over ferroalloy bars under flux AN-348-A, the burn-off of alloying elements reaches the following magnitude: for chromium—2.5%, for vanadium—2.1%. Molybdenum and tungsten practically do not oxidize. The burn-off of elements is determined on the basis of chemical analysis of the slag crust.

To determine the effect of the degree of nonuniformity in the dosing of ferroalloys in the bar on the chemical composition of the deposited metal along its length, we carried out the following experiments.

From the above-mentioned mixture of ferroalloys, bars of the following dimensions were made: \(6 \times 16 \times 50\) mm (weighing 26 g), \(6 \times 16 \times 100\) mm (weighing 52 g), and \(10 \times 16\) mm of variable height (weighing \(1.2 \div 9.2\) g).

Bars of variable height and different weight made it possible to “assemble” a bar of dimensions \(6 \times 16 \times 50\) mm (weighing 26 g) from five bars, each weighing 3.2; 4.2; 5.2; 6.2; 7.2 g and 1.2; 3.2; 5.2; 7.2; 9.2 g, with a bar base area of \(10 \times 16\) mm. Assembly of a \(6 \times 16 \times 100\) mm bar (weighing 52 g) was provided by joining separate bars weighing 3.2; 3.7; 4.2; 5.2; 5.7; 6.2; 6.7 and 7.2, as well as 1.2; 2.2; 3.2; 4.2; 5.2; 5.2; 6.2; 7.2; 8.2; 9.2 g.

The individual bars were laid tightly on the surface of the workpiece over a length of 350 mm. The scheme of the “assembly” variants for different dosing of ferroalloys is shown in Figs. 1, 2, 3.

Surfacing was carried out on workpieces measuring \(18 \times 18 \times 370\) mm. To prevent spreading of the deposited metal, graphite molds were used.

Fig. 1

The graphite mold projecting 3–4 mm above the workpiece ensures good formation of the deposited metal and prevents arc burning between the electrode and the graphite. The electrode was short-circuited at a distance of 1 mm (in the clear) from the ferroalloy bar in the middle of the workpiece. Completion of the surfacing operation (stopping the feed of the electrode wire) was carried out 66 sec after the “Start” button was pressed. Dosing of the arc-burning time ensured the obtainment

of the same chemical composition of the deposited metal both in the middle and at the end of the bead.

Comparison of the bead composition (by chemical analysis) \(b_5\) and \(b_{10}\) with the presumed composition, according to the dosing of ferroalloys \(a_5\), \(a_{10}\) (Figs. 2 and 3), shows—

Fig. 2

Fig. 3

—shows a special property of the weld pool. This property consists in the fact that the composition of the bead is equalized along its length despite the nonuniform dosing of ferroalloys. The weld pool combines regions with different concentrations of elements into a “mixing reservoir.” This property of the weld pool may be called its “leveling capacity.” The leveling capacity of the weld pool can be explained by the fact that, within the limits of the weld pool, the chemical composition of the metal is the same throughout the entire volume, independently of the uniformity of the linear dosing of ferro-

alloys. To verify this assumption, the arc was broken at arbitrary positions along the length of the workpiece, on the surface of which nonuniform dosing had been carried out according to the scheme \(3.2 \div 7.2\) and \(1.2 \div 9.2\) g/cm. The places where chips were taken and the results of the analysis are indicated in Table 2.
The deposited metal was not annealed before the chips were taken¹.

On the basis of this property of the pool, calculations were made of the concentrations of elements in the pool when dosing ferroalloys at \(1.2 \div 9.2\) and \(3.2 \div 7.2\) g/cm, with bar lengths of 5 and 10 cm. The calculation results and experimental data are given only for tungsten (Figs. 1, 2, and 3).

It is known that any periodic function (in this case a broken straight line) can be represented by a series of sinusoids—the fundamental and higher ones. The higher harmonics are “cut off” by the leveling property of the weld pool. Graphical integration of the function expressing the assumed composition (according to the dosing of ferroalloys along the length of the bead (\(v_5\) and \(v_{10}\); Figs. 1, 2, and 3)) gives the averaged distribution of elements along the length of the bead.

The calculation was carried out for such conditions as provided a weld-pool length of 70 mm.

The discrepancy between the composition assumed by calculation (\(v_5\) and \(v_{10}\); Figs. 1, 2) and the actual composition is explained by the fact that the graphical calculation was made on the assumption that the full pool length (70 mm) is established immediately at the beginning of the bead. In reality, after arc initiation the pool reaches its full size only after a rear portion 30–40 mm long has solidified. This phenomenon is difficult to take into account in the calculation, since the growth of the pool dimensions in the unsteady period has not yet been studied. Later, however, the effect of this section weakens, and the curves of the calculated (assumed \(v_5\) and \(v_{10}\)) and actual chemical composition (\(\delta_5\) and \(\delta_{10}\)) coincide.

The periodic alternation of dosing fluctuations (Figs. 1, 2, 3) is the worst case for leveling the chemical composition of the bead along its length. Other arbitrary combinations of dosing irregularities improve the leveling.

Experiments showed that the greatest leveling of composition is achieved when using bars of a length equal to \(0.7 \div 1.3\) of the weld-pool length.

Such a bar length ensures a close location of the greatest and smallest values of the dosing irregularities. A shorter bar length increases the cost of their manufacture and does not improve the leveling property of the weld pool.

The analytical calculation² of the bead composition as a function of the ferroalloy dosing scheme along the length was carried out as follows (Fig. 4).

Let the dosing of ferroalloys vary according to the equation:

\[ Y_{\text{dosing}} = a - bx. \]

To simplify the calculation, we take the length of the weld pool \(l_v = 1\); \(l_v\) is taken as the unit of measurement for the total surfacing length as well.

When the pool moves to the right by \(dx\), the change in concentration \(dy\) in it will be expressed as:

\[ \frac{Y_k + Y_{\text{dos}} - Y\,dx}{l_v} - Y = dy, \]
where

\(Y_k\) is the intermediate concentration of the pool,

\(Y\) is the concentration established in the pool after its displacement by \(dx\).

Hence

\[ \frac{dy}{dx} + Y - Y_{\text{dos}} = \frac{dy}{dx} + Y + bx - a = 0. \]

The solution of the first-order linear equation will be:

\[ Y = C \cdot e^{-x} - b(x - 1) + a. \]

In Fig. 4: \(Y_{\text{dos}} = a - bx\), “equation for dosing ferroalloys”; \(x_0 = l_v = 1\), \(y_0\), \(y_k\), \(dy\), \(l_v = 1\), \(dx\), \(dx\), \(dy\), \(y_k\), \(y\).

Fig. 4

We determine the constant \(C\) from the initial conditions (Fig. 4):

\[ \text{at } x_0 = 1,\quad y_0 = a - \frac{b}{2}, \quad C = -\frac{b}{2}\,e. \]

Then

\[ Y = -\frac{b}{2} e^{-(x-1)} - b(x - 1) + a. \]

Example. The length of the surfacing deposit is \(L = 21\) cm (Fig. 5).
The length of the weld pool is 70 mm. The assumed content of the element according to the dosing is 20 at the beginning of surfacing and 5% at the end. The change in concentration follows a straight line.

Then

\[ b = \frac{20 - 5}{3} = 5,\qquad Y = 20 - \frac{5}{2} = 17.5. \]

The concentration of the element at the end of the bead at \(x = 3\) is

\[ Y = -\frac{5}{2} e^{-(3-1)} - 5(3 - 1) + 20 = 9.66. \]

The initial difference in the composition of the surfacing deposit at its ends, \(20 - 5 = 15\%\), decreased to \(17.5 - 9.66 = 7.8\%\).

Figure 5 gives the graphical calculation for this case. The results of the graphical and analytical calculations coincide. According to studies in large-scale metallurgy, the rate of distribution of alloying additions in liquid steel is sufficiently high. In the presence of additional stirring of the pool, this occurs extremely rapidly. Precisely under surfacing conditions we have a small pool weight (80–120 g), overheating of the metal, and its stirring due to the presence of oscillations and convective

currents. All this leads to a rapid equalization of the chemical composition of the bead along its length.

Experiments show that as a criterion of inadmissible nonuniformity of the composition of the bead along its length one may adopt a technological factor—the presence of lack of fusion.

Fig. 5

lack of fusion. In Fig. 3, \(a_5\) and \(a_{10}\) are shown—the tungsten content in the deposited metal assumed according to the ferroalloy dosing. According to the scheme of ferroalloy dosing without taking into account the leveling action of the weld pool, the fluctuation in tungsten content reaches \(2.6 \div 17.5\%\); \(b_5\) and \(b_{10}\) are the calculated tungsten content in the deposited metal taking into account the leveling action of the weld pool. The fluctuations in tungsten content decrease and amount, for a bar length of \(5\ \text{cm}\), to \(b_5 = 9.5 \div 12\%\); for a bar length of \(10\ \text{cm}\), \(b_{10} = 7 \div 11.5\%\). However, with dosing according to the scheme \(1.2 \div 9.2\ \text{g/cm}\), lack of fusion is observed in the dosing section \(9.2 \div 9.2\ \text{g/cm}\). Lack of fusion appears as a result of the fact that the amount of ferroalloys introduced through the bar exceeds the fraction of base metal being melted. Lack of fusion in this case may serve as an indication of inadmissible nonuniformity of ferroalloy dosing, although in individual cases fluctuations in tungsten content of \(9.5 \div 12\%\) may be considered permissible for deposited steels of type R9.

Analysis of the deposited metal over the height of the sections shows the absence of zonal segregation (Table 3). This makes it possible to regard the deposited metal, over the entire section, as equivalent to the cutting part of the tool.

Equalization of the chemical composition of the pool apparently cannot occur without equalization of its temperature. The average temperature of the pool [3], \(1700 \div 1900^\circ\), extends over the entire length, with the exception of the active spots of the arc and the crystallization zone. Equilibrium [8], in our opinion, should more appropriately be related to the average temperature of the pool, which “holds” for \(8 \div 14\ \text{s}\) (with a weld-pool weight of \(80 \div 120\ \text{g}\)), than to the solidification temperature of the steel. The duration of stay of individual regions

of the deposited metal in the interval of crystallization temperatures and, therefore, is considerably less than the holding time at the average temperature.

The experimental data of K. V. Lyubavskii [7] prove that the oxygen content in the deposited metal during automatic welding of low-carbon steel coincides with the equilibrium content if the equilibrium is referred to the average temperature of the pool—1800°.

Table 2

¹⁾ The length of the weld pool reaches 65–70 mm. Samples at distances of 20, 40, and 60 mm from the end of the bead are within the contour of the “last pool.”

Table 3

Chemical composition over the height of the section of the deposited metal, %

Note: The height of the deposited layer reached 7–8 mm. The layer-by-layer sample—taking chips for chemical analysis—was taken every 1.5 mm, starting from the fusion zone, over the height of the deposited layer.

Conclusions

The leveling property of the weld pool has been investigated. This property consists in the practically instantaneous leveling of the composition of the liquid metal in the volume of the weld pool, despite the nonuniform dosing of ferroalloys along the length of the weld. Analytical and graphical calculations of the chemical composition of the pool have been proposed for determining the permissible nonuniformity in the dosing of ferroalloys. In individual cases, a technological sign—lack of fusion—may be taken as an impermissible nonuniformity of the chemical composition of the deposited metal along the length.

Lack of fusion occurs when the quantity of ferroalloys in the bar exceeds the share of base metal in the deposited metal. When bars are used whose weight per unit length is more than 80% of the weight per unit length of the deposited metal, the latter is formed only at the expense of the bar and the electrode metal. The power of the welding arc in this case is insuffi-

sufficient for melting not only the flux, electrode metal, and ferroalloy bar, but also the base metal.

The crystallization characteristics of the deposited metal lead to the elimination of zonal segregation. The entire deposited metal constitutes an equivalent cutting part of the tool.

REFERENCES

1. Alov A. A.—On the mechanism of crystallization of weld metal. Avtogennoe delo, No. 9, 1949.

2. Pogodin-Alekseev G. I. Theory of welding processes. Mashgiz, 1950.

3. Khrenov K. K.—Ceramic unmelted fluxes for automatic arc welding. Collection of papers of the scientific-technical conference of welders. Mashgiz, 1953.

4. Frumin I. I.—Alloying of deposited metal in submerged-arc welding. Avtomaticheskaya svarka, No. 1 (22), 1952.

5. Brinberg I. L.—Modern equipment for automatic submerged-arc welding. TsNIITMASH, Book 14. Problems in the theory of welding processes. Mashgiz, 1948.

6. Pokhodnya I. K.—Automatic surfacing of a layer of high-chromium wear-resistant steel of type Kh12 on backing steel. Avtomaticheskaya svarka, No. 4, 1953.

7. Lyubavskii K. V. Metallurgy of automatic submerged-arc welding. Problems in the theory of welding processes. TsNIITMASH, Book 14, Mashgiz, 1948.

8. Lapidus V. A.—Study of the equilibrium state of the metal–slag system in welding with electrodes of grade TsI-1. Avtogennoe delo, No. 4, 1951.

9. Kosmachev I. G.—Automatic surfacing of multi-blade tools. Mashgiz, 1952.

10. Automatic electric-arc welding. Mashgiz, 1953.

1. The change in composition at the end of the pool is explained as follows: when the “Stop 1” button is pressed, the feed of the electrode wire and the movement of the tractor cease. The welding current is not switched off. Burning of the arc until self-interruption additionally melts the base and electrode metal without corresponding melting of the bar of ferroalloys, since the tractor is stationary. When the “Stop 1” and “Stop 2” buttons are pressed simultaneously, the pool has a constant composition. ↩↩

2. In carrying out the analytical calculation, the author used the consultation of Associate Professor V. A. Nadezhnikov. ↩