Haptic augmented reality (AR) is an emerging research area, which targets the modulation of haptic properties of real objects by means of virtual feedback. In our research, we explore the feasibility of using this technology for medical training systems. As a possible demonstration example, we currently examine the use of augmentation in the context of breast tumor palpation. The key idea in our prototype system is to augment the real feedback of a silicone breast mock-up with simulated forces stemming from virtual tumors. In this paper, we introduce and evaluate the underlying algorithm to provide these force augmentations. This includes a method for the identification of the contact dynamics model via measurements on real sample objects. The performance of our augmentation is examined quantitatively as well as in a user study. Initial results show that the haptic feedback of indenting a real silicone tumor with a rod can be approximated reasonably well with our algorithm. The advantage of such an augmentation approach over physical training models is the ability to create a nearly infinite variety of palpable findings.