Preheat - the mysteries revealed

TWI Bulletin, July/August 1991

Neville Gregory's career as a metallurgist spanned 40 years up to his retirement in 1990. After working at The English Electric Company, Rugby, Murex Welding Processes and Enfield Rolling Mills he joined the London office of BWRA as a Welding Technologist. For six years he carried out liaison visits to members and dealt with technical enquiries.

This work continued when he joined the Research Station at Abington and was combined with contract research on various welding processes and materials including zinc coated steels, low alloy steels, reinforcing bars, and armour plate. Since 1975 Neville Gregory headed a Welding Advisory Service in the Arc Welding Department assisted by a team of Welding Engineers.

Efficient application and selection of preheat can prevent hydrogen cracking in welded joints and contraction cracking in hardfacing deposits. As Neville Gregory explains there is more to this than just deciding on a minimum preheating temperature.

The most common form of cracking in a welded joint is hydrogen cracking in the heat affected zone (HAZ). It can also occur in the weld metal, particularly when the strength of the weld is high enough.

Hardfacing deposits are less prone to HAZ cracking because the restraint is generally much lower than in a welded joint and cracking usually takes the form of transverse cracks in the weld caused by shrinkage stresses.

The usual precaution to prevent occurrence of such defects is to preheat the parent metal before welding. This is well known, and because preheating is common practice there is some surprise when it does not appear to work. A study of case histories over many years shows that even though preheat temperatures have been correctly selected, if their application is not carefully controlled there can be a reduction in quality that may require expensive rectification.

The principles underlying the correct application of preheating to prevent cracking in various materials are described below.

Carbon-manganese steels

These steels include those specified in BS 4360 and BS EN 10 025, and also ASTM A36, A242, A373, and A441. Recommendations for preheat temperatures are given in BS 5135 and AWS D.1.1-90. Determination of preheat temperatures by the method in BS 5135 takes into account the chemical composition of the steel, its thickness, joint type, the energy input of the welding process, and the potential hydrogen content of the weld metal.

Experience has shown that BS 5135 provides effective estimates of preheat temperatures, but whatever source of data is used for it to be effective, these guidelines should be followed:

1. The steel composition should be checked whenever possible by spectrographic analysis, even when mill sheets are available.

   The importance of this will be appreciated when it is considered that a difference in carbon content of 0.05% can change the preheat required by 50°C. In thick plate, i.e. above 50mm, this represents a considerable difference in energy costs in a large fabrication. This, of course, is insignificant compared with the cost of repairs if the preheat is 50°C too low.

2. As recommended by BS 5135, when preheat is applied locally to the joint the required temperature should be reached in the parent metal for a distance of at least 75mm in any direction from the joint preparation. If local preheat is applied by gas flames, whether manually or by fixed burners, the temperature of the component should be checked on the side remote from the heating source. If this is not possible the temperature should be measured on the heated side of the joint some time after removal of the heat source.

   This is to avoid measuring the high surface temperature immediately after removal of a gas flame. A delay of two minutes for each 25mm of parent metal thickness is recommended.

3. The preheat to be used during fabrication should be applied before tack welding, and the tack welds should be at least as large as the root runs to be deposited. This is obvious if hydrogen cracking associated with tack welds is to be avoided, but it does represent the ideal world. In many cases tack welding is carried out without any preheat and the tack welds and their HAZs (possibly cracked) are ground out during fabrication. Alternatively, a deep penetration process such as submerged-arc welding can melt out the tack welds and their HAZs. Experience has shown that both methods are effective but it would be a wise precaution to check this by procedure testing.

   Another procedure is to use large tack welds with sufficiently high heat input to avoid the need for preheat. This procedure should be checked with reference to BS 5135. It is important to note that when preheat is required for tackwelding, the required temperature should be reached at least 75mm from the position of the tack weld.

4. Most methods for determination of preheat give minimum temperatures for avoidance of HAZ cracking. These temperatures may have to be increased if restraint is particularly high, for example when welding a large complex structure. In cases of high restraint, even though the preheat may be high enough to avoid HAZ cracking, there may be weld metal hydrogen cracking. In many cases no harm will occur if the preheat is increased by, say, 75°C and the maximum interpass temperature rises by 200°C or more above the calculated preheat.

However, an increase in interpass temperature may be accompanied by a decrease in weld metal strength because softer microstructures are formed by the lower cooling rate. For example, undiluted weld metal from basic coated electrodes may suffer a decrease in yield strength of 75 N/mm ² if the interpass temperature is increased from 150 to 350°C.

A limitation on the interpass temperature may also be required when maximum fracture toughness is required in the HAZ. In all cases the maximum interpass temperature should be included in a welding procedure specification and it is vital that this interpass temperature should be strictly adhered to during fabrication. This can increase the total welding time, possibly double it, with a corresponding increase in costs which are frequently underestimated because of this hidden factor.

In critical applications, when optimum weld metal and HAZ quality and mechanical properties are required in heavily restrained joints, it is evident that welding procedure testing must be carried out, and the results scrupulously applied during fabrication.

Quenched and tempered steels

If preheat or interpass temperatures exceed certain limits it is possible to reduce the specified tensile and yield strengths and also the toughness of the HAZ.

The steel makers and some codes of practice specify minimum and maximum preheat and interpass temperatures for different energy inputs and plate thicknesses.

These recommendations should be strictly adhered to and the extended welding times should be allowed for when estimating costs.

Medium carbon and low alloy steels

As the carbon and/or alloy content of a steel increases so does its hardenability, so that hard HAZs are formed more readily, i.e. at lower cooling rates than in carbon-manganese and other structural steels.

Preheating temperatures are generally higher, covering a range from 150-300°C, and because of the increased tendency to hardening of the HAZ, even in relatively thin sections, the preheat temperatures do not usually vary with section thickness.

However, as explained below, sometimes it may be advantageous to increase the preheat temperature as the section thickness increases to facilitate hydrogen diffusion from the joint.

For successful welding of low alloy steels a preheat is chosen that is high enough to prevent hydrogen cracking, even though the HAZ may be partially transformed to martensite at this temperature.

At high temperatures, e.g. 1250°C, the HAZ is transformed to austenite which on cooling starts to transform to martensite at a particular temperature depending on the composition of the steel. On cooling further, the HAZ progressively transforms to 100% martensite through a temperature range of 100°C or more. For example, a particular batch of AISI 4340 (BS 970 817M40 or En 24) containing 0.4C, 2Ni, 1Cr, 0.3Mo starts to transform to martensite when cooled to 290°C, and when the temperature falls to 250°C the HAZ contains 50% martensite and 50% austenite. Upon further cooling more martensite is formed until at 200°C the HAZ has completely transformed.

Experience has shown that if the above steel is preheated to 250°C and welded with low hydrogen consumables, cracking does not occur provided that certain other conditions are met. In welded joints made under low restraint, e.g. a butt weld between two plates free to move, ensuring slow cooling after welding by covering with insulating blankets or burying in insulating powder has proved to be a satisfactory procedure for many years. However, when restraint on the joint was higher cracking sometimes occurred and the solution generally used was to increase the preheat temperature to 300°C or higher.

If the temperature was high enough cracking was prevented for two possible reasons:

1. A proportion of softer transformation products occurred in the HAZ with decreased crack susceptibility.
2. Hydrogen diffused away from the joint region.

The above procedure is somewhat hit or miss and today two safer alternative procedures that ensure freedom from cracking are used:

1. After preheating and welding the component is allowed to cool to a temperature at which martensite transformation in the HAZ is complete but above that at which the steel is susceptible to hydrogen cracking. It is then heated to550-650°C to temper the hardened HAZ.

   The Table shows typical preheating and pre-tempering temperatures for some common alloy steels.

2. If the above procedure is not convenient, an alternative is to heat the component immediately after welding to 250°C (which for some of the above steels is the preheat temperature) and to maintain this temperature for long enough to allow some diffusion of hydrogen from the joint. A duration of four hours is generally sufficient but this period may have to be extended by many hours for heavy sections. The component is then cooled to room temperature. The HAZ will be sound but have low fracture toughness and in many cases it is advisable to temper the component.

Preheat and postheat temperatures

A third alternative procedure is used in special cases, generally for small components that can readily be placed in a furnace. The procedure is to preheat to a temperature above that at which the austenite in the HAZ starts to transform to martensite, but at a temperature that allows transformation to a softer microstructure that will not crack. This temperature is maintained during welding and for some time afterwards by placing the component in a furnace at the appropriate temperature for sufficient time for complete transformation of the HAZ to a soft microstructure. For the 4340 steel a suitable procedure would be to preheat to 350°C and hold at this temperature for two hours.

For this procedure temperature control must be precise to, say ± 20°C. Postheating 4340 to 400°C for two hours would give only 50% transformation to soft microstructures and subsequent cooling would transform the remaining austenite to martensite. Although cracking may not be a problem, because some hydrogen would diffuse out of the joint at the high temperatures involved, the toughness of the HAZ would be reduced which might be significant depending on the service conditions of the component.

For maximum HAZ toughness a procedure should be used that produces 100% martensite in the HAZ which is then tempered.

It will be evident from the comments above that to obtain the most benefit from preheating the quality of a welded joint must be considered in terms of mechanical properties as well as freedom from cracking. This requires careful control of preheat, interpass temperature and post-weld heat treatment.

The time-honoured method of increasing the preheat temperature until cracking no longer occurs is unfortunately still in use but it can cause serious deterioration in the mechanical properties of both the weld metal and the HAZ, and as described by Hart TWI Bulletin 1974 15 (12) (375-376) is not always a satisfactory solution.

Hardfacing alloys

The commonest form of cracking in hardfacing alloys is contraction cracking caused by the low ductility of the weld metal which is unable to accommodate the shrinkage strains that occur during cooling.

Differential contraction between weld metal and parent metal can be reduced by preheating, and the higher the preheat the lower are the contraction stresses.

Therefore in hardfacing, as opposed to welding, there is some justification for the philosophy of increasing the preheat until cracking is prevented. Hardfacing alloys having increasing hardnesses above about 400HV require progressively higher preheats from 150 to 400°C.

Apart from hardness, geometrical factors have some influence on the preheat required. For example, mechanised hardfacing processes such as submerged-arc, mechanised MIG, TIG, flux-cored arc or plasma transferred arc (PTA) produce deposits having smoother surfaces with finer ripples than manual hardfacing. The coarser ripples of manual deposits can act as stress raisers which accentuate any cracking tendency so that a higher preheat is necessary.

Geometrical factors affect crack susceptibility when medium carbon or low alloy steel cylinders are hardfaced. Weld metal applied to the outside of the cylinder has a higher cracking tendency than a deposit on the internal surface. Therefore the preheat required for hardfacing the outer surface may be higher than that for the internal surface if a crack sensitive alloy is to be deposited.

For crack sensitive alloys a general rather than a local preheat is more effective because it reduces hot spots and results in a lower cooling rate.

Interpass temperatures may have to be restricted to 100°C above the preheat temperature and slow cooling after welding may be required, making it necessary to wrap the workpiece in insulating blankets or to bury it in vermiculite granules.

Components of complex shape may have variable section thicknesses which cool at different rates. This can cause cracking and it may be necessary to control the cooling rate by placing the component in a furnace previously heated to the relevant preheat temperature.

In critical cases post-weld heat treatment at 600-650°C may be required to stress relieve the component, and heating and cooling rates should be a maximum of 50°C/hour.

Conclusions

Cracking occurs in welds and HAZs for a number of reasons which may be simple or which may encompass many interacting factors.

This article and the two previous - Why do welds crack? ( Bulletin 1991 March/April issue) and Cracking of hard facing alloys (May/June issue) - have given some guidelines on how to prevent cracking and therefore produce welded fabrications to the required quality within the quoted price.

It must be emphasised that welding and hardfacing procedure specifications should be fully documented and should be based on realistic tests or relevant experience.

This on its own is still not sufficient to guarantee success. Apart from specifying the procedures correctly it is essential that they are applied and controlled by adequate supervision.