Typically component bodies are made from Non Wetting materials. Solder bridges between leads of a device or across devices can usually be avoided though PCB pad design and in the case of leaded devices – by changing their clinch direction so the distance between the offending leads is increased.

Solder will not attach to non-metallic surfaces. Therefore the component body need not be masked. On the other hand solder bridge between component pins is a definite possibility, which can only be alleviated by proper component layout or orientation. 

Also as a general design rule try to place SMT ICs on the top side of the PCB so that it does not have to be soldered using wave process. If the design does not allow this then specialized fixtures or masking becomes necessary to avoid any bridging of the leads.

It is not normal for the solder to stick to the SMT components. This sounds like there is not enough flux on the board when it exits the wave.

This could also be the cause of the bridging on the SOIC. Even with proper flux you can still get bridging on SOIC's if you do not use solder thieves.

In wavesoldering any termination that will wet to the solder while being submerged has the ability to bridge. It's more a function of the lead size and spacing that causes the actual bridge. The closer they get the more bridging can occur.

As for solder adhering to plastic. If the SMT bodies are going to contact the solder wave i would suggest you either use a selective soldering pallet to hide the SMT for the wave. If your using a bottomside glue process and the SMT will be and can be solder by the wave, make sure you flux all areas coming in contact with the wave. If not oxides (not solder) can adhere to the components.

The common solution is using solder pallets or solder fixtures. This is a sheet of high temperature composite material about .500 inches thick that is machined to match the shape of the PC board with "pockets" machined in where the SMT components are and holes machined through where you want solder to go.

These "pockets" protect the components from touching the solder. These act as a carrier to carry the PC board down the line though the wave solder machine. These work very well to solve the problem you described. If you google solder pallets you will find several companies that make them.

Securing components on the bottom side of the printed circuit assembly is a common practice. The chip wave was developed to address this issue of surface mount component on the bottom side of the board. The orientation of the chip components relative to the direction of travel had become an issue and if they were oriented with the termination perpendicular to the wave then the potential of skips (i.e. missing solder) was very high.

This is where the chip wave came into play, its function was to splash solder at these difficult locations creating a pathway for the larger wave to complete the solder joint. When SOIC were assembled to the bottom side of the board, they also needed to be placed perpendicular to the wave so the solder could flow from one lead to the next as opposed to if the component was oriented parallel to the wave the wave would only see two set of leads, the front and the back side. This orientation was susceptible to many short between the leads and it was all due to the ability or the inability of the solder to flow off the component leads.

In some designs an extra pad was added to each side of the component to allow the solder to flow off the last lead. Therefore orientation was most important to prevent solder shorts and reduce the amount of solder skips.

Another important issue was the profile used to solder these components. Regardless of the soldering process used to create the solder joint, the thermal ramp up through the preheat still had to be 2 to 3 degrees C per second and the thermal shock could not exceed 80 to 100 degrees C and as can be imagine this was quite a task considering the components were going to be subject to the molten wave which was typically set at 250 degrees C.

Solder getting stuck on the component body has not been observed to be an issue as the component bodies are not metallic and the solder will not wet to non metallic surfaces.

Another technique is to use fixtures to prevent the solder from touching all the surface mount components. In this case however the surface mount components have to be installed using the traditional method of solder paste, pick and place and thermal reflow processes. After this surface mount process the through hole components can be installed and the board can be placed in a fixture which just exposes the areas to be soldered around the plated through holes. The fixture is then place on the wave solder system and the process of soldering the through hole component is completed.

There is plenty of information available regarding component orientation and wave solder considerations by doing the appropriate searches on the web.

If more information is needed, please contact me.