Hello Tommyjoking, distortion control is part science, part controllable and uncontrollable variables, and part intuition. You are certainly right when you described the scenario in your post. Not everyone can attain the "knack" for solving distortion related issues and it isn't something that can readily be explained either. Over time many individuals are able to recognize and put into action processes and steps for limiting and controlling distortion problems, some aren't able to recognize or achieve this. Many of the variables that I have run into in this regard are related to how the parts were produced, what sorts of internal stresses are present in the parts to be fitted and welded, which welding process is utilized for welding these parts, this can include things such as different wire diameters, shielding gases used, and in the case of GMAW, which transfer mode is involved. All of the different welding processes can lend or detract from the ease with which distortion can be controlled in a given situation. Individual welders habits and machine setting comfort zones can also have an effect on controlling distortion. Things such as presetting angles to amounts that are slightly greater than the finish angle will allow for shrink to be controlled and help with keeping certain types of parts square or at the correct angle. Heatsink materials will also aid with controlling distortion and warp in some instances. Tightly jigging a part and restraining it while welding can help in many instances as it limits the ability of the parts to move and at the same time may allow the jig to act as a heatsink as well. Setting a bow in certain parts prior to welding and allowing the part to cool before releasing it will also aid with keeping a part straight if that is of importance, in this instance it might require some experimentation with setting the amount of bow, before the correct amount will result in a flat finished part. Through experimentation, careful observation and lots of practice/experience many will be able to judge how to proceed with welding out a weldment to minimize the effects of welding distortion or heat related distortion.
     Extruded and rolled shapes can react very differently to heat and welding processes, in some instances application of heat to these items can actually release internal stresses that will warp or distort them in ways that typically might not follow the general rules of heat and shrinkage(I was taught that generally when heat is applied to a part it will expand and as it cools it will shrink slightly more than it expanded). In the case of some rolled or extruded shapes I have experienced the release of internal forming stresses that override the heat/shrink rule and thus cause the shape to distort in a manner that might not be expected. I think this somewhat explains the intuitive part of recognizing these sorts of things ahead of time and being able to compensate for this. You and the other contributor to this thread have brought up a great topic that has many answers and considerations for all of us out there who are faced with issues for fabricating better parts, pieces, etc. Regards, Allan