I have a deterministic system response with two continuous factors that can be varied in any continuous combination with each other, A1 and A2. I have a third continuous factor, B, that only affects the system response if a third categorical 'switch', C, is True. So:

C == False: A1, A2 only affect system response

C == True: A1, A2, and B affect system response

In the past, we have typically made two DoE sampling designs to handle the C == True and C == False cases separately. But is there a more efficient way to handle this in one DoE? Should I be combining factors somehow--perhaps B and C in some fashion? I am interested in an interactive GUI solution if possible.