A conventional implant/bone construct typically involves a plate fixed to a fracture bone via set of screws aligned vertically to the plate/bone interface. In this study new configurations of plate/bone constructs with varied screw angulations were tested. The conventional screw configurations a parallel set of screws aligned vertical to interface were tested against two new configurations, called the divergent configuration and the convergent configuration. A total of four construct systems were prepared for tests. The samples used in the experiments are listed. Sample A, named conventional construct, used a set of parallel screws drilled vertical to the interface plane. Sample B, named convergent construct, used screw holes drilled such that their axes converged. Samples C and D, named divergent construct, used screw holes with diverging axes. The difference between Samples C and D is in the angle of alignment between their screw axes and the normal to the interfacial plane, the former drilled at 150 and the latter at 300. The load displacement curves for the four samples are shown. The failure loads, along with the mode of failure for each sample are read from the load-displacement curves. At a first glance, it is observed that the sample with highest pullout strength in terms of failure load is the conventional construct (A). This is followed by the divergent constructs (C and D), and the sample with the lowest strength is the convergent construct. It takes a more careful investigation, however, to note that similar results will not apply when one takes the average pullout stiffness as a criterion. Here, we define the average pullout stiffness of the structure as It is immediately noticed that samples producing the highest stiffness are those prepared as divergent constructs (C and D), followed by the convergent construct (B), and finally, Sample A produced the lowest stiffness. Also note that the initial portion of the load-displacement curve for Sample D (300 screws) is highly nonlinear, although the curve then assumes a linear shape. The stiffness obtained in the linear region of Sample D, and that of Sample C are quite close in comparison.

This initial deviation in the curve of 300 divergent screws may be explained by a self-aligning motion of the screws as the load is applied, during which the angle of alignment tends to decrease towards a smaller angle. As the angle gets closer to that of Sample C and the curve becomes linear, the stiffness produced by Sample D, approaches to that of Sample C.

As a result, we have found the divergent constructs to be a promising alternative to the conventional set of screws, used in treating fracture involving bones with osteoporosis. Moreover, research needs to be focused on determination of an optimal angle of alignment, the number of screws to be used and the optimal distance between each screw for the divergent construct.