I must admit that I still do not quite get the picture. Are you welding around the circumference of the flanges of the head and tubesheet, instead of bolting up the flanges? (This is not good engineering practice.)

Also, what materials are being welded together? If you were using 309, it would suggest that you are welding C/Steel to S/Steel. Is this correct?

Without having the full picture, here are a couple of comments:

1)If the cracking problem that you were describing in the root was indeed "hot cracking", then you have a contamination problem. Hot cracking can be seen by the fact that it forms while the weld is still hot. Typically the welder will see the crack forming while he is still welding. It also occurs along the centreline of the weld. (Also sometimes referred to as centreline cracking.)
2)Hot cracking is not as a result of heat absorption by the base metal. It is caused by contamination of low melting point constituents, typically S or P or Pb.
3)Hot cracking is a typical problem in austenitic weld metals.
4)To prevent this, you can either get rid of the contamination, or us a filler that will result in at least 5% ferrite. Your 29/9 fillers (E312) are very good at eliminating hot cracking. (It gives around 40% ferrite)
5)Some dye penetrants contain sulphur, and could be the source of your contamination. When performing DPI, it must be removed completely before welding again.
6)Pre-heating may remove some volatile contaminants, but appart from this, pre-heating of austenitic materials is not normally of any great benefit. It may reduce residual stresses if you can ensure a very even heat distribution, otherwise, it does not even give you this benefit.
7)Pre-heating austenitic materials will typically result in excessive grain growth of the weld metal and heat affected zones of the austenitic parent metal. This results in reduced toughness and strength. It can also result in sensitization of the HAZ if the parent metals are austenitic stainless steels.

I assume that what you mean by "tensile boundary" is some residual stress forming between the weld metal and parent metal. From your description, it sounds as though the weldment was subjected to a post weld heat treatment. (PWHT) Again, this is a problem for austenitic materials. Your typical austenitic stainless steels will sensitize if this is not done correctly. If the one material in the joint is a ferritic material (e.g. C/Steel) and the other is an austenitic material, then the PWHT will not typically result in a great reduction in residual stresses because of the difference in co-efficients of thermal expansion.

From my experience with hot cracking problems, it can be very insiduous. Sometimes the problem is seen in the root of the weld and the offending weld metal is removed by grinding and re-welded. Following this rewelding, it looks like the cracking has been eliminated, but there is still some cracking deeper in the weld. The following passes tend to propogate these cracks through the thickness of the weld due to the great thermal stresses experienced during the welding.

Personally, I would suggest that you perform weld build-ups on the flange faces to compensate for the distortion that you described. Then machine the flange faces down to the required thicknesses and bolt up the joint as it was designed originally.

Let us have some more info regarding the materials used in the joint, and the PWHT cycle that you used. It may give us further clues as to what may be wrong.

Hope this helps

Regards
Niekie Jooste