In the CFA (complete Freund's adjuvant)-induced Inflammation model that we used, we observed activation of mononuclear phagocytes, maturation of dendritic cells and release of TNF-a, IL-12, IL-6.
What is it that you are looking for? Because there are a lot of inflammatory agents that will get you all the observed effect of CFA, but it won't be the same mechanism.
If you just want to use the components of CFA, it's MTB, i think it's the Ra strain.
If you want to mimic the contribution of the oil component and how that affects the MTB cell wall lipids.. there are whole studies on the use lipid coated beads.
However, CFA in vivo is almost always in an emulsion with another immunogen, and it's mostly used for vaccine responses... And it's a good antibody inducer.. so I'm not sure exactly what part of CFA you are trying to mimic in vitro.
Unless you have a BSL-3, using virulent MTB is out of the question.
Ra is a non-pathogenic stain, I think it's rated BSL-2.
If you know someone who works with MTB, I would start with MTBRa, or you can technically also use M.bovis BCG pasteur strain will work also if you can't get Ra. Without the oil component, live bacteria is always better. And all MTB stains (even the non-virulnet ones) are very good inducers of Th1, which is what you are aiming for.
As for the type of cells... depends... if you want to look at maturation, you can start with (in mice bone marrow derived or in humans PBMC derived) differentiated macs and Dcs. It'll also work on all sorts of monocytic/macrophage cell lines. For MTB, you almost always need macs or DCs since they are the cellular targets. For protocol, I would suggest you look up papers using MTB Ra for in vitro infection. Or even MTB or BCG. This is a very well researched area.