Scale models of Triceratops prorsus skulls (primarily based upon the holotype specimen YPM 1822) were used to test horn locking positions. The models were produced as part of the "Favorite Collection" skull models line, sculpted by Takashi Oda for Kinto and Company. The models were cast in poly-resin, and the skull and lower jaws were integrated into a single unit.

The models were sculpted at approximately 15% natural size. In order to verify their accuracy, they were compared with photographs, drawings, and measurements of YPM 1822. No major differences were noted between the scale model and the original specimen (see Appendix). Models were used in this study rather than full-size casts because the small models were easier to manipulate; it is presumed that this would achieve results similar to those found using full-size replicas. Also, models are appropriate here because scale-dependent factors (such as intracranial stress and strain) were not tested.

It was assumed that skull morphology accurately represents the shape of the head in life. However, a keratinous sheath may have increased horn length by a variable amount (Hatcher et al. 1907). Thus, the methodology presented here tested "horncore locking" rather than "horn locking," but the latter term will be used for simplicity. For the purposes of this study, it was assumed that the length of the horncores approximated the length of the horns (i.e., horncore plus horn sheath). Deviations from this assumption are addressed in the discussion.

Two identical skull models were used concurrently to determine positions in which the horns could interlock. Multiple positions were tried, with two constraints. First, the horns could not intersect with cranial fenestrae or foramina, such as the external nares or supratemporal fenestrae. In life, such action would have been harmful or fatal to the animals and not conducive to sustained horn locking. Thus, it is presumed that they would have avoided this behavior during wrestling matches (however, this assumption does not consider the possibility of a well-placed horn thrust used to disable an opponent and end the wrestling match). Second, the skull was constrained to rotate down 90 degrees or less from horizontal (as defined by the maxillary tooth row), reflecting limits in neck movement. This degree of movement is somewhat arbitrary, as the mechanics of ceratopsid neck motion are largely unexplored. Ultimately, this last constraint proved inconsequential, as it was found that most horn locking positions took place when the skulls were at an angle of 45 degrees or less from horizontal.

The skull models were photographed in all possible horn locking positions, and the relative skull positions were noted for later analysis and interpretation.