I'd say you were on the right track with your original approach, however it is possible that there was insufficient pre heat, excessive joint restraint or the cooling rate was too fast.
Let's start at the beginning. Cast steel is primarily a reference to the method of forming the component and tells you nothing about the possible composition of the component other than informing you that it is steel, rather than iron and thus implying a reasonable degree of weldability. A photo of the component would be handy when you get sufficient posts up to be able to upload one.
I don't know your experience level, so forgive me if I state the seemingly obvious.
In order to perform the more critical forms of welding such as this, you will need some method of accurately measuring the pre heat and interpass temperatures. Temperature crayons or non contact infra red thermometers are the two most common methods that spring to mind. Without knowing the exact composition of the component, I'll make a bit of an assumption and work on quite a high Carbon Equivalent (a measure of the weldability of steels). Based on this you could be looking at preheat temperatures of anywhere from 170 deg C for a 10mm material thickness up to 260 deg C if the component is 40mm thick. For a highly restrained joint (one unable to freely expand and contract while welding and cooling), you would possibly add another 50 deg C to those temperatures. Generally, too much pre heat is better than too little. Electrodes should be used from a hot box to ensure they are dead dry and truly Low Hydrogen. Your TIG root run is optional, but make sure you achieve 100% penetration so you don't allow a crack to propagate from an unfused root. It may not hurt to use a higher classification electrode than a 7016 such as 8018 or even 9018, but based on what you have said, I believe your issues are more procedure related than consumable related. As far as pre heat goes, a down and dirty method for gauging temps is to remember that at 100 deg C, water will quickly evaporate from the surface (nearly instantly), while at 200 deg C the drop of water will dance around on the surface in a very energetic fashion as it boils off. You will want to heat a generous area so that the cold metal doesn't rapidly suck the pre heat away from the weld zone. It doesn't hurt to check the temperature during longer welding jobs and adding heat as required as 3.2 electrodes don't put much heat in and therefore heat loss may be greater than heat input. If you can run them, some 4mm sticks may be a better option. I would definitely grind each run back to shiny metal so as to eliminate potential inclusions and slag traps and would also peen each run as you progress as well as the final capping pass. A cheap air chisel with the edge blunted is a really good tool for this otherwise hammering will achieve the same result. Once you are satisfied that all welding is complete then you will want to cool the component slowly and evenly to eliminate thermal shock and resultant stresses. A fiberglass welders blanket is ideal to pack around the joint so as to exclude wind and slow the cooling rate.
I recommend you download this book, some 212 pages long (only 2.7Mb though) that apart from listing the various Welding Industries of Australia consumables, has a large technical section detailing various steel grades and the procedures and consumables to weld them.
http://www.welding.com.au/news/view/australian-welding-guide-available-to-download
Hope this helps, but having limited information to go on, there are many unknown factors.