Is the designer's concern that the parts will fall off when hanging upside down during reflow? We've been doing this for decades and our test if we think a part is marginal is to glue a 2nd (and sometimes 3rd) part to the back and send it through reflow. If the part doesn't fall off with double its weight then it'll be fine. When we first started doing this we had a formula that used part weight, wettable area, and surface tension to see if the part would stay on, but these days we can pretty much eyeball it and run an experiment with the ones we're not sure of. All but the heaviest parts will be fine.
You also need to be concerned about moisture sensitivity. You need a system for controlling the time between 1st and 2nd reflow if you have moisture sensitive parts.
Lastly, make sure none of the parts have restrictions on the number of heat cycles. Probably not an issue, but just to be thorough count up the number of heat cycles the board will see (eg. 2 reflows, 1 possible repair, selective wave, etc...).
reply »