RAM Concept - Load History Deflections Not Converging

Question

Hi, 
 I have two RAM Concept suspended slab models which will not converge when running the load history deflections and I am at a loss as too why. Below is the warning provided in the Load History Calc Log: 
 
 WARNING: An error has been found while calculating load history deflections. The load history deflection solution did not converge. The expected deflection error is 1.803 mm 
 
 Information which may be helpful: 
 
 Suspended slab minimum thickness 600mm 
 Models of Basements 4 and 5 - Walls all around with internal columns 
 Models of Ground to Basement 3 are very similar and do converge 
 I am unsure if this is purely coincidental but the columns change from rectangular in G-B3, to circular in B4 and B5.

Calc Option Parameters are in screen shots 
 
 Changes made to try and resolve issue 
 
 Iterations to use increased to 20 (both under zero-tension and load history) 
 Checked with user only reo and with program reo 
 Fixed meshing warning of some wall centrelines not joining 
 Models of Ground to Basement 3 have the same warnings and do converge

Are you able to provide any guidance on how I may be able to get these models to converge in order to get accurate load history deflection results?

Seth Guthrie · Answer

It would be best to have the model, then we can see your load steps, the supports, first order deflection, calc logs etc. . Secure File Upload 
 There are some useful tips in the program manual in the Load History chapter that might help. It could be due to runaway cracking, for example.

Seth Guthrie · Answer

In your file, MEL MSCP L3 PT Design_Rev5_Alt.cpt, gives a warning in the 9th iteration of the "Sustained Load" step that's fairly small. 
 WARNING: An error has been found while calculating load history deflections. The load history deflection solution did not converge. The expected deflection error is 1.099 mm 
 Since the peak deflection for this step is over 30 mm, that seems like a minor deviation. 
 There are a number of sections that are significantly cracking in the analysis, but the overall deflection is about 3x the first order service deflection which is quite reasonable using a creep factor of 3.3 and a duration of over 10,000 days. 
 Using wide band beams with no torsional stiffness may be a factor. You could try rerunning it with "Normal" 2-way slab behavior on the beams to confirm.