In surgery procedures, haptic interaction provides surgeons with indispensable information to accurately locate the surgery target. This is especially critical when visual feedback cannot provide sufficient information and tactile interrogation, such as palpating some region of tissue, is required to locate a specific underlying tumor. However, in most current surgery simulators, the haptic interaction model is usually simplified into a contact sphere or rod model, leaving arbitrarily shaped intersection haptic feedback between target tissue and surgery instrument less unreliable. In this paper, a novel haptic feedback algorithm is introduced for generating the feedback forces in surgery simulations. The proposed algorithm initially employs three Layered Depth Images (LDI) to sample the 3D objects in X, Y and Z directions. A secondary analysis scans through two sampled meshes and detects their penetration volume. Based on the principle that interaction force should minimize the penetration volume, the haptic feedback force is derived directly. Additionally, a post-processing technique is developed to render distinct physical tissue properties across different interaction areas. The proposed approach does not require any pre-processing and is applicable for both rigid and deformable objects.