welding cast iron to structural steel

huangyhg

welding cast iron to structural steel
i was asked to look at a remodling project where the "architect" called for a new steel beam supporting second floor wood joist to be jacked into place and sit atop existing cast iron columns.  literally sit.  
the round ornamental cast iron columns are approximately 14' tall.  a steel plate is laid on top of the columns, then the new steel beams sit on the plate.  the simple span beams meet over the plate, but there is no connection between beams, from the beams to the plate or the plate to the cast iron column.  every thing is sitting loose, with only the vertical dead weight of the upper floor "holding" the framing together.
my question is what welding rod can be used to weld the cast iron to the structural steel (a36 or equivalent)?
check out our whitepaper library.
you'll have some problems welding to cast iron.  expansion from the molten metal will probably crack the cast iron, for one.  i think you'd have to preheat the cast iron first, as a minimum.  cast iron is brittle, and does not like tensile forces.  try bolting if possible, although drilling through cast iron is also difficult.
it's hard to get a good idea of the geometry from your description.  but can you put a collar projecting downward a bit from the plate?  this would be to make sure the column didn't move horizontally relative to the beams.
welding to ci is a problem, as mentioned above.  but perhaps the bigger problem is that you won't really know what kind of welds you have when you're done.  this would be the case when welding to old steel, as well.  you can't qualify a weld when you're welding to "mystery metal".
lppe & jstephen:
thanks for the quick response. i was thinking along the same lines as your notes indicate, but i was wondering if anyone had actually made a welded connection like this.
i'm pretty sure the c.i. columns are in the 100 year old range.
between the new steel beams and the c.i. columns is a loose 1-1/4" thick steel cap plate.  i just don't feel comfortable leaving the beams sit loose on the columns as they are.  i'll have to give some thought to devising a collar for the columns and perhaps bolt up to the loose plate on top.  then weld the plate to the beam.
thanks again.
i have done a project involving a cast iron bridge. the eic told me the problem with old cast iron is that the outer layer anctually inhibits corrosion. once this layer is damamaged or removed, the cast iron becomes highly suseptible to corrosion.
you need to check if the column is truly cast iron or wrought iron, being 100 years old i suspect it might be wrought iron. simple field metallography techniques can be used. wrought iron have a wood grain looking microstructure, long stringer of silicates, surrounded by pure iron.
if it is wrought iron then clean and weld with an austentic stainless steel, like 308. though don't expect it to have reliable mechanical properties. that is to say, the high silicon slag that contained within the wrought iron microstructure could cause inclusions in the weld deposit.
if it is cast iron, find some way to bolt it together.
vita sine litteris mors est.
i wouldn't try welding unless you know what the column is and know you can replace it.
welded repairs to cast iron are iffy at best and it sounds like you would have a heck of a time heating a column that size to a high enough temp to prevent weld tear out and cracking upon cooling.
i ran a precast plant that made manholes (vertical pipe), it was a zero slump process that used vibration under pressure to force a ci bell ring form into the concrete.  these machined rings were made of cast iron (not cast malable steel).  when the rings cracked my welder, ground into the ring and used nickel rods to for welding.  preheating was difficult but was absolutly esential.  the work was limited to short beads to prevent distortion and more preheat.  five hours of repair time would buy anywhere from two uses of the ring to fifty uses.  it was not worth the effort.
a steel (and welable) cap that would cup the ci column and be mechanically restrained seems to offer a connection that may be analized and thus make a more dependable connection.  my three and a half cents (sense).