It also helps to remove moisture from the weld region, in addition to the above. Removing moisture is critical to avoid having hydrogen or delayed cracking in certain steels.
The following is applicable to carbon and low-alloy steels:
The faster the cooling rate, the greater the tendency to form harder microstructures, in the following progression: ferrite+pearlite; bainite; martensite.
Higher temperature in the surrounding base metal will slow down the cooling rate and increase the time (T800-500 time is the usual measure) and avoid harder structures that are more prone to hydrogen cracking. The peak hardness is usually in the coarse-grained heat-affected zone, because peak temperatures and cooling rates are maximum there. Greater heat sink effect is why thicker plates require higher preheat (all else being equal).
Alloy steels require greater preheat (all else being equal) because the alloy elements tend to promote harder structures more rapidly; this property is 'hardenability'.
Preheat cannot exactly remove moisture* (the unavoidable source of hydrogen), but permits H to diffuse more rapidly away from the weld region, so it is likely to be at lower concentration in the HAZ after things cool down. H absorption occurs at high temperature, but the cracking occurs at ~room temp, hence the term 'delayed cracking'.
* Moisture is never actually present in the steel as such, but it easier to use that term than to explain concepts like dissociated hydrogen and diffusion rates.
I've had people tell me that steel absorbs moisture, which you can see coming out when you heat it with a torch. I don't usually bother to enlighten them so they don't get embarrassed.
As a young welder I was told preheat was used to drive moisture out of the pores in the steel and to reduce thermal shock. Based on those concepts no one thought there was a problem with "driving" the moisture out of the steel and then go off to take a coffee break or lunch.
It wasn't until I went back to Ohio State that I learned the real reason or the real need for preheat and why it was necessary for steel alloys, but not for aluminum alloys.
Slowly, but surely, the AWS CWI program has helped to put some of the old wive's tales to rest. Things like storing low hydrogen covered electrodes in gas fired cooking ranges, shorting the low hydrogen electrode against the work piece to "dry" the electrodes, and others are being debunked by the CWIs on the job site and in the fab shops. It is a task without an end in sight, but still worth the effort.
