Picture attached

You are probably getting an initial agglomeration reaction with the acid, water, and ore.  While chains might help for a while, since the agglomeration is starting there you will probably continue to have build up, eventually also on any chains or flaps.  Only positive way is to increase the clean out frequency.

from the look of your agglomerator the water addition is too high=> very sticky materials and build up. this is due to the moisture addition being at or even slightly above the material max stickiness - the sticky limit. less water addition will be the prime reason for the build up.
two issues to check
1. the water addition is the first one to hit the ore. the agglomerate size is set by the water/ore ratio and occurs within one or two revolutions. the acid should be added to the agglomerated ore and not to the un-agglomerated ore. the acid heat of dilution on the outside of the agglomerates helps to indurate the agglomerate. also the conc acid will dissolve some calacium silicates that then re-precipitate on the outside of the agglomerate forming a stronger agglomerate. the precipitated gypsum and silicate will also have much lower stickiness with the majority of the clay contained within the agglomerate.
there is a major issue in controlling the water addition with changes to:
- the ore feed moisture content
- the ore mineralogy esp clays, silicates, etc
- the tonnage rate feed to the drum.
all of these need to be measured and the moisture addition rate changed to suit. we included an "Split-On-Line" PSD monitor that also measures the reflectence of the agglomerates.
the agglomerate PSD decreases in fines with more moisture addition. BUT the agglomerate refelcence is a good indicator if the moisture addition is too high. this was coupled with on over belt moisture meter on the feed conveyor so that the dry tonnes could be calculated and the water addition adjusted to suit the incoming % H2O.

I have a couple of papers that i have written on designing, operating and controlling agglomerators. hapy to send copies if requested.
and yes only floppy rubber liners will assist with accretion removal.

relying in operator control of the water and acid addition does not work very well as they are generally quite busy and do not really have the background knowledge to control an agglomerator. it is one of the simplest machines but one of the hardest to operate consistently well without including all the measurements and control mentioned.