As Terry already said the way to go are Mathcads numeric methods, either a solve block or the "root" command.
The problem with Mathcads symbolics is twofold:
1) The symbolics does not know anything about units and treats them as unknowns (which also may be negative or non-real). Considering this, no solution is found. Sometimes it helps to tell MC that a unit is real and positive by adding "assume,m>0" to the symbolic eval, but in your case it does not help. What works is to tell MC that m=1 (as seen in the screenshot below. Thats bad as you lose Mathcads unit checking feature, you may get wrong results as of unit discrepancy and you would have to add the correct unit yourselft afterwards. And ... you may get a highly undesirable result as seen in the picture, because
2) when the symbolics is forced to float mode because no exact solution is possible, it will not give you all possible results of your equation but only a single one and usually there is no way to control which one (adding "assume,B>0" or the like has no effect.
So the symbolics is of no use in your case and thankfully isn't necessary at all. You should have set up your sheet to make all variables which depend on B functions of B. I was too lazy to change your sheet that way and so I used the symbolics to make the expression delta.c-delta.a1 a function of B which you want to be zero.
The sheet shows how to use "root" to find a solution. Guess its the same 3 mm which Terry found using a solve block.
