Convergence Issue and change material during analysis

[TPdct]

My model has several zerolength element with ElasticNoTension(ENT) Material and several ShellNL element with Drucker Prager Material.
It fails to converge at the first step. When I change the DruckPrager Material to Elastic nd material, it works fine and the result is reasonable.
Can I change the material during analysis? At the first few analysis steps, when the load is small and unable to make the material yield, use Elastic Isotropic Material to get the right result of stresses and strains, and in the subsequent steps when the load is larger, change the material to Druck Prager Material and use the result of elastic material as initial guess. Is there anyway to do this?

[fmk]

not w/o you writing some c++ code