by Chuck Nicodemus

Money Saving Method of Torch Work on Steel Members

Reprinted 2005

The use of the oxyacetylene flame to do and assist in doing many difficult, and some apparently impossible tasks in steel work is often underestimated. There are a large number of jobs that beginners as well as older welders "pass up", because they have not become thoroughly acquainted with the possibilities presented by the use of contraction, resulting from heat properly applied.
To get the best results it is necessary to study very carefully the action which takes place in a piece of steel when the heat of the flame is properly applied and the metal left to cool.

I will try to give you a brief description, and drawings of a few jobs, hoping to assist you as you study and apply this time and money saving method of torch work on steel members.

This "contraction method" will always help, and in many instances completely do a job of straightening bent steel members or bent steel members to fabricate something otherwise difficult. I have used this method to straighten power shovel booms, heavy structural members, steel bridge portals, wind braces and cords with out removal from original

Other parts have been straightened, truck-frames, truck-body sides, end-gates, sills, cross members and countless other steel articles, to say nothing of having used in steel fabrication and repair work. This process is applicable in all steel shapes. So far it has scarcely been investigated by the average welder. Yet he should find it of great value whether he is repairing or fabricating anything of steel.

We will try to confine this discussion to the regular steel shapes as milled and fabricated in different structures, and thus get some knowledge of how to apply it to straighten or bend whatever may come our way whether large or small.

All welds or heats cause or relieve stresses, and often change the structural characteristics of the metal. A little careful study will aid any operator in using the heat to a great application. Then, too, he will soon find himself doing many well paying jobs, which before seemed quite impossible, although he may have only a good oxy-acetylene torch and gases.

The electric welder also should take this method into consideration if he is going to get the best results from his efforts, for a weld improperly applied may cause him much trouble if the stresses are wrong and are not relieved.

For the beginner we will start with simple experiments and try to develop this method to a wise and workable art, step by step. It is often stated that steel contracts on cooling as much as it expanded on being heated. This is true in volume but not in alignment or length, if the heat is properly applied. Now it is up to us to apply the heat in the proper way to straighten, bend, or shorten the steel member to meet our needs.

When we speak of heat pulling the steel we must remember that the heat properly applied only forces the metal to upset, or expand into itself, and that the contraction as the metal cools is the force which pulls it towards the desired point. Then, too, each heat on cooling will pull only a certain distance, therefore, if a great amount of change is required in the alignment, the work can be hastened and the cost of the same lowered by adding some outside force while the metal is hot.

Steel expands or contracts a definite amount for each degree of change in temperature, but we shall be reasonably accurate if we figure this expansion or contraction to be on eight of an inch per foot of heated metal if such metal be heated to the ordinary temperature used by the blacksmith. The method of applying the heat must be such that the steel instead of expanding in length will upset, or expand into itself, so to speak. Also to make this method work there must be portions of the member cold enough, strong enough, and so situated as to force the metal to upset, or expand into itself, when and where heated unless some outside force can be added.

To begin, take a flat bar about one quarter of an inch thick, by two inches wide, by approximately two feet in length. Select a fair sized tip and begin to heat this bar about one half an inch from edge A (as in Fig. 1) with the tip, or flame pointed slightly towards edge B. Hold the flame steady until this point becomes heated to a light red but do not melt the surface at any time.

When this spot C in Fig. 1 and Fig. 2 has reached a light red heat commence to wider the heat as you move the torch slowly toward the other edge B, and finish the heat at the edge B (as shown in Fig. 2). Let A be the side from which you start to heat, C the point at which you start, and points C,D, and E show the part of the bar heated. After this specimen has cooled, note how much it has pulled edgewise toward the widest part of the heat.

This is what happens: The metal at A and B is cold, but at C the flame heats it rapidly and forces it to expand. As A and B are rigid the metal at C shoves together, or upsets, or expands into itself, as some would express it. This continues as the heat moves toward B. only to a greater extent as the heated section is widened. Also as the flame moves from the point C the metal at that point drops in temperature, therefore it becomes more ridge and does not restretch, but as it contracts as it drops in temperature, assists in upsetting the hotter metal at the torch tip.

The widened heat and the contracting metal behind it upset the edge B, which does not restretch but contracts on cooling so that the bar is bent edgewise, the amount of bend being dependent upon the width of V heated. The point A should not be heated more than the temperature created by the conductivity of the metal, or in other words the cooler the point A, the better. This point acts as a hinge and the V shape of the heat allows an equalization of stresses to a large extent.

After this bar has cooled and its change in alignment noted, reverse the procedure, heating in a V form from B to A and the bar on cooling will become straight if both heats are equal, but you will find that the bar has shortened considerably. In this manner you can shorten and thus lighten bars and members without removing them, even on such structures as bridges.

Now take a piece of angle iron, say one quarter of an inch in thickness with legs two inches wide. This angle to be approximately two feet long. Leave one leg cold but starting at point A (Fig. 3) heat the other leg in the same manner as you did the bar in Fig. 1 and Fig. 2 and you will get the same results. If you wish to straighten this angle, reverse the heating procedure as was done on the bar, and when the Leg which was left cold the first time is reached, heat rapidly a narrow strip across it and widen until the heat on this leg of the angle is as wide as the widest part of the heat on the other leg of the angle (See Fig. 4). As before the zig zag lines show the path of the torch.

To describe more fully: In Fig. 3, let A be the starting point of the heat and B the finish. This draws the ends toward the edge B and is the easiest angle bend to make.

In Fig. 4, the heat starts at C and widens as the lines show as the torch moves toward the side A. Then the upright leg of the angle is heated as indicated by the lines. This pulls the ends of the angle toward the side A as the metal cools. The angle of Fig. 3, can also be straightened by stretching the portion which had been heated. This is done by placing it on a heavy piece of steel and striking the part which had been heated a few blow with a hammer, or by holding a sledge on one side and using the hammer on the other.

This stretches the metal and acts, or accomplishes just the opposite from the contraction caused by heating the metal properly and letting it cool. You will profit by remembering this operation, as many times it will be of great value in connection with a heat on the opposite side or edge of a member to be bent or straightened, as may be desired.

Now let us try a piece of channel iron. Heat it as we did the bar and angle only on both flanges. Leave the webb cold and starting from that side of the flanges widen the pattern as we approach the edges, as was done in the previous experiments. (See Fig. 5 ) The cooling flanges contract and bend the channel away from the webb.

To bend this channel toward the webb start the heats in from the edge of the flanges a short distance and as they progress toward the webb, widen them in a V as in previous experiments. When the webb side is reached , heat the webb as you did the leg of the angle in Fig. 4. (See also Fig. 6). The heats begin at B and progress toward A.

To bend a channel edgewise (see Fig. 7) Let A represent one flange, B the webb or back and C the other flange. Leave the flange A cold. Start to heat the webb B at the base of A and widen the heat as you approach the flange C then heat the flange C its entire width to a length equal to the widest part of the heat on the webb. To change the direction of the bend simply reverse the heating procedure.

To bend the I beam sidewise, proceed the same as on a channel iron except that the webb of the I beam is heated to the same width as the heats on the flanges a the points where they join the webb. (See Fig. 8).

To bend an I beam edgewise heat it the same as the channel iron in Fig. 7. LEAVE the flanges cold on one edge of the webb, heat the webb in a V shape starting close to these flanges, and widen the heat as you move toward the other flanges using the same procedure as in previous experiments. When these flanges have been reached heat them so that the entire heat is in the form of a V from start to finish. (See Fig. 9)

On a job requiring a heat like any of these experiments, if a greater bend or alignment change is desired than can be obtained by one heat it is advisable to add some outside force to assist in upsetting the metal while it is still hot. If it is impossible to add some such force, let the member cool and then re-heat it, repeating the process until the desired bend or alignment change is obtained.

It must be kept clearly in mind that the entire heat, through the whole section or member which is to be bent or straightened, must be in a complete V form from one side or edge to the other, whether it is of a single piece of made up or a number of pieces fabricated into one member, so that on cooling the contraction will also be in the V form and thus, to a large extent, eliminate internal stresses.

If you have carefully noted the action of each piece of steel in these experiments you will have seen that the widest part of the heat in the bar or member, has necessarily been on the which was to be shortened.
If the heat is to be applied to straighten a member that has been bent, the widest part of the heat must be on the point or outside of the bend. These experiments must be made to get a thorough and sure knowledge of the exact action which will take place when the metal cools after being properly heated. This must be studied very carefully.

At times when an operator is applying a heat, there occurs a warping or increase in the amount of bend. This is caused from heat conducted and radiated to the adjacent metal and need not be considered. Although , after a little practice the operator will find it possible to use the expansion for what is, what it is, to an advantage as he becomes better acquainted with this method of bending and straightening steel.

All the beginner needs to take into consideration is the metal which is heated enough to upset, unless there is an excessive weight or stress on the member being heated. If there is excessive weight or stress on a member which is being heated it must be relived if it be of an amount sufficient to stretch the metal while it is hot. Ordinarily this is unnecessary as there usually is enough rigid metal to support considerable weight or stress.

You may sometimes find that a heat, though properly applied, does not pull as much as you think it should. There may exist internal stresses which the first heat may only relieve and a second heat will be required to do the pulling. Then too, you will find many bends that will respond to what seems but very little heat. Practice alone can fit you to be a good judge of these conditions.

The better grades of steel are often heat treated. If you encounter such a member be sure to keep in mind the fact that heat "anneals" or softens it. If this annealing weakens it for its particular use or purpose, so reinforcement should be added to offset this weakness. If a bar, angle, or any other steel shape is added for reinforcement, weld on the edges only. Never weld crosswise of a member unless absolutely necessary or to weld a crack.

Ordinary members of mild steel need no reinforcement unless subjected to excessive stresses. On such structures as bridges, which are always subjected to excessive vibration or reverse stresses, it is advisable to reinforce. Here too, never make a weld crosswise on any member unless absolutely necessary.

Photo No 1, Page 12, (not shown in this report) shows a heavy 10 inch I beam which was used as a boom. It was bent while in service and looked like a very difficult task for a small shop to attempt to do. It was straightened with nothing but the oxy-acetylene heat applied properly to accomplish the task. This I beam responded very readily to the contraction method, and not once was a hammer or any force used. Other than its own weight setting as shown and contraction caused from heat properly applied to upset the metal in the right places. This I beam was not moved from its position during the process. This eliminated handling cost which alone would have been more than the cost of the gases used, if handled, as is necessary in a small shop. The time required to heat it was no more than would have been necessary to make a single heat in a forge and at least three heats would have been required. The torch heat was applied to the exact spot desired, thus avoiding all short kinks, and physical exertion.

One thing must be remembered: Do not over heat, since in a case like this where the weight of the member is assisting the heat, the change in alignment takes place very rapidly and it is easier to reheat than to pull the member back. No place should be heated to such an extent that the weight of the member will start it to sag, but let the contraction pull it to the desired place if you want a prefect job when finished.

"An added note"
In my iron workings, I used an air hose with a water pick-up to cool the hot spots. This enabled faster work and let me start more hot spots on very large bridge beams, up to 300 pounds per foot. A person can use a bucket of water and rags to soak the hot spots also, but it is rather messy.

Good Luck!

To contact the author of this article email: Chuck Nicodemus

Return to Home Page

Created: 12/28/2005
Last Rev: