(2) The maintenance welder must have a great many more talents than a production welder.  First of all, in production welding it is usually the case that the base metals being worked on are clean, new metals.  This is not so in maintenance. Often the maintenance welder is faced with salvaging equipment which may be many, many years old, having had service in corrosive conditions, may be oily or greasy and so dirty and contaminated that everything in the text book goes wrong when the welder attempts to weld repair it.

(3) In production welding, it is usually possible to position the work so that the welding can be done in a convenient position, usually down hand.  This is not so in maintenance welding because as often as not the maintenance welder must repair objects and in awkward positions which he can hardly see or reach, let alone weld.

(4) An additional difficulty in maintenance welding is the wide variety of work that must be accomplished.  Often in production welding, an operator will work on a limited number of jobs constantly.  In maintenance welding, the operator does not do one type of work constantly; and as a result, he understandably cannot become proficient in every type of work that he does, because certain types of breakdown occur so rarely.  It is extremely difficult for a mechanic to learn how to do all the myriad projects required in maintenance welding efficiently.  One of the most difficult problems in maintenance welding is that the welder often does not know the analysis of the base metal.

(5) In spite of the fact that maintenance welding is more complex than production welding, it is undoubtedly true that maintenance welding is far more profitable to a plant or industry to do than production welding.  Frequently, a maintenance welder does in one day, work which may save his company hundreds, if not thousands of dollars.  Such savings are not possible with one man's time in one day in production welding.

(6) The first step in maintenance welding is to determine the base metal.  Knowing something about each metal will help identify metals.  Spark tests, hardness tests, magnet tests, chemical tests, weight tests and file tests are common methods of identifying base metals.  However, there are often cases where it is almost impossible to be certain enough for safety by common shop methods of analysis.  In those cases, it is imperative to use a welding filler metal with the highest physical properties to make certain that the weld equals or exceeds the base metal irrespective of what the base metal may be.

The second step in establishing a welding procedure is to calculate the effect of the heat to be applied.  All welding requires heat, and heat will cause a certain reaction to the base metal.

The heat generated in a weld is predictable from the formula OH=A2RTO (Heat equals amperage squared times resistance times welding time).

The undesirable effects of heat can be listed as excessive grain growth, hardening cracks, porosity, thermal cracks, warpage, locked-up stresses, distortion, and hydrogen contamination.

(7) The non-uniform localized heating and cooling during welding and the joining of the heated base metal by means of the molten weld-filler metal creates a hindrance to both expansion and contraction.  The stresses arising through heating and cooling of the base metal are called contraction or shrinkage stresses.  The stress system left in the object being joined after welding, due to thermal or shrinkage stress, is called the residual stress.

(9) As in any metal casting into a mold, stress in the weld metal resulting from hindered contraction is related in intensity to the dimensions of the weld.  Therefore, the maximum stress is in the direction of welding, longitudinally.  The transverse stress is next intense and the stress in the thickness direction is least because less hindrance to contraction occurs here.

(10) Welds contract in all three directions - length, breadth and width - and the resulting stress may be called multi-axial stresses.  In maintenance, welders are constantly called upon to solve welding problems where multi-axial stresses are a source of anxiety.  The thermal stress problem is accelerated when heat is applied locally and is dissipated into the base metal mass.  The harmful results of stress are both complex and of serious concern in maintenance welding.

(11) The temperature gradient is the heat-affected zone; that is, the area starting from the centre of the weld to the extremity to which the weld heat travels.  Within this heat affected zone most welding problems are created.  Some sections of this heat affected zone may be cooling while other parts are still being heated, which contributes to the thermal stress problem.  Unless there is an equal amount of residual compressive strength in the metal system to balance the residual tensile strength, cracking will occur.

(12) The problem created by stress and distortion causes several difficulties.  First, they restrict normal ductility of the material.  Second, they may cause localized stress corrosion cracking that may fail under impact load.  Stresses may exceed the yield strength of the base metal and result in cracking.  Additionally, a loss of dimensional stability occurs through distortion.

The amount of stress and distortion which occurs in a part being welded depends upon a number of variables such as thickness of plate, degree of restraint, speed of electrode travel, movement of air, preheating, higher heat input and other factors.  It is generally assumed by most engineers, however, that a weld will contract approximately 3mm for each 2.5cm of weld across section transversely.  Longitudinally, a weld will, in general, contract or shrink approximately 2-3mm for each 3m of weld length.  The expansion and contraction rate of metal produces serious internal stresses and only requires a slight excess strain to exceed the yield strength of the metal and produce weld failure.

(13) Another serious problem in maintenance welding is that of a martensitic zone adjacent to a weld.  When hardenable steel and cast iron are heated into their critical range and allowed to cool faster than their critical cooling rate, a brittle martensitic zone tends to occur next to the weld.  This is due to the limited graphite rejection in the region adjacent to the frontier zone between weld and base metal.  Other problems which occur in this region are carbide precipitation, grain growth, porosity and hardening graphite's. If a martensitic zone is allowed to occur.

(14) In addition to the problems already mentioned that occur in maintenance welding, an added problem is that of stress raisers.  Any factor which produces a localized area of high stress is called a stress raiser.  Any engineer is aware that abrupt changes in section design, notches, grooves, screw threads, surface irregularities and discontinuities such as cracks, holes and inclusions, are considered stress raisers. However, in maintenance welding, we are only concerned with those avoidable notches which occur as a result of welding.  These notches have very little effect on the tensile strength of ductile materials but are of great importance in fatigue.  The notch sensitivity factor depends not only on the material but on the type of notch and level of stress.  Those notches which are avoidable are crater cracks, hard spots, undercuts and porosity.

(15) Take, for example, a typical butt weld.  There are three starting points for fatigue fracture.  These are; internal defects an undercut at that point where the weld makes a junction with the plat or base metal; and poor quality of weld at the root.

(16) The shape of the welding bead has a considerable influence on stress raisers, especially on cast iron and the hardenable steels.  For example, if a weld bead is applied to a cold piece of base metal, at the beginning of the weld, the weld will appear to be convex and lap over at the cold start.  This makes perfect stress raiser and as such it will be highly efficient in starting a crack.  Additionally, when the electrode is abruptly removed from a weldment, there will be a crater at the end of the weld.  A crater is often a source of cracks because a crater solidifies from the outside towards the centre.  Since the weld crater is a smaller mass than the remainder of the welding bead, it will cool at a faster rate than the heavier section.  These conditions usually result in a starter crack and the creation of a severe stress raiser.

When a martensitic zone, residual stress or distortion results after a weld has been made, these conditions can be improved by stress relief or mechanical relief.  However, the only practical solution is to anticipate these problems before the welding is accomplished and to apply corrective measures to avoid their occurrence during welding.

Some of the techniques which we have employed to eliminate or minimize stress and distortion follow.  None of these techniques are empirical or can be used in every case, nor are any of them absolutely foolproof. In may cases it will require more than one of these corrective measures because in some instances, one alone will not be sufficient.

(3) The next step is to join the two sections together using substantial weld bead to prevent a crack.  By reducing the cross section of the weld area substantially, we have greatly reduced the tendency for contraction and thus we will have less stress and less distortion.

(6) The anchoring technique, when working on dirty, oil saturated cast iron removes contaminated metal and exposes the subsurface sound metal.  Secondly, we are anchoring the weld metal into the base metal in much the same way that a snow tread tire gives better traction than a smooth tire when operating in the snow.  However, most important of all, we have broken up the continuity of a vulnerable martensitic hardened zone adjacent to the weld.  Thus, when stresses are applied, rather than the weld failing adjacent to the weld, the continuity has been broken up so the strain will not be focused at one vulnerable zone.  Additionally, the grooves create a mechanical bond and also result in more metal-to-metal contact for greater holding power.

(7) The anchoring technique is of immense value before applying hard facing alloys to heavy equipment and is especially important when welding cast iron.  We have seen jobs accomplished successfully in this manner which were attempted time and time again with failure with other methods.

(8) One if the most important ways to control stress and distortion is the practice of peening, which consists of tapping the weld bead while still not with a rounded tool (such as a ball-peen hammer).  The reason for peening is that when a warm weld bead is peened the weld metal is stretched and expanded.  This stretching of the weld bead compensates, at least to some extent, for the contraction which will occur upon cooling.

(9) There are several important things to know about peening. It is standard practice to peen all but the first and last pass.  If you are peening upon and air-hardening tool peened, cracking may occur.  Therefore, the first pass should not be peened. Subsequent passes should all be peened up to the last pass, the cover pass.  The reason these are not peened is that a peened weld, and this is true even if it is mild steel, is a work-hardened weld bead, and a work-hardened weld is an efficient crack starter.

(10) The internal passes will not be work-hardened because the subsequent weld beads which are applied over them will anneal the work-hardened condition and does not cause cracking.

(11) Incidentally, stress relieving after welding does not always relieve peening damage, but subsequent welding does.  Therefore, the rule in peening is to peen all but the first and last passes. it is important to use moderate blows because repeated moderate blows are much better for peening than a few heavy blows.  It is imperative that the peening tool be light in weight and blunt rather than sharp in design.

(13) One of the most universally used methods of controlling distortion and stresses is that of preheating.  Preheating before welding eliminates or lessens the danger of crack formation, minimized hard zones adjacent to the welds, minimizes shrinkage stresses, lessens distortion and enhances the diffusion of hydrogen from the steel.  A rough but realistic rule of thumb is that a 260 ^oC) preheat usually equals 800 ^oC of post heat (just as an ounce of prevention is said to equal a pound of cure).