Product Lifecycle Management (PLM) provides abundant tools to define products, combines information related to the products, and exchanges such information between different actors in their lifecycles. Among these tools, Computer Aided Engineering (CAE) has the advantages to create and edit a product in a digital format and may supply numerical simulation methods for analyzing functionalities of a product. The main drawback of such simulations lies in two aspects. On one hand, the implementation of a CAE simulation, e.g., Finite Element Analysis (FEA), tends to be a time-consuming process. On the other hand, the interaction modes during these simulations remain relatively poor. For example, engineers cannot access to intermediary simulation results in order to adjust simulation parameters in an interactive way. This drawback decreases the efficiency of information flow and inclines to influence the process of the PLM in a negative way.In other areas, development of information technology boosts the invention of new solutions based on advanced technical equipment that bring the human operators closer to the scientific data. Virtual Reality (VR) is such a promising domain in which an operator is immersed in the product space characterized by realistic renderings, multi-sensory, and intuitive interactions. Thus, VR technology unseals a terrace with a large variety of potential applications, ranging from massive scientific data explorations, surgical trainings, to virtual prototyping.Design evaluation, within an industrial context, of deformable mock-ups in a VR environment could take benefit of the introduction of the human operators into the loop.Moreover an interactive design validation of such mechanical parts plays an important role in a PLM, because an interactive deformation simulation in a VR environment allows engineers from different industrial sectors to immerse themself and to manipulate these digital mock-ups for the purpose of identifying design problems in the early design phase.The time and costs required for sharing product information among different sectors would so be largely reduced, and therefore the efficiency of the design information exchange in a PLM could be considerably increased.However, interactive deformation simulation in a VR environment through haptic interfaces is a challenging task caused by the trade-off issue between the deformation accuracy and the real-time interaction performance. This issue would be troublesome for simulation cases in which complex mechanical deformable parts are taken into account. Namely, the degree of deformation realism tends to require a high meshing resolution, which has a direct impact on the size of the matrices involved in the elasticity system and thus brings about a huge computational task in real-time.Our two-stage deformation simulation method for real-time haptic interaction purpose that combines an off-line pre-computation phase and an on-line deformation interaction phase may help to tackle these difficulties. This chapter will illustrate the action of virtual reality with haptic on the PLM loop by introducing the user in the loop.We propose a deformation evaluation framework for real-time haptic interactions by introduce a two-stage method: an off-line phase to pre-compute deformation spaces, similar to a model reduction method but based on modal analysis and an on-line phase to enable haptic interactions by a costless response model. We propose a mesh analysis method for the pre-computations during the off-line phase. This method allows the off-line phase to calculate different deformation spaces with an accuracy enhancement regarding correspondent anticipated scenarios. Moreover the method permits a real-time switch among the different pre-computed deformation spaces, so that the on-line deformation computations focus on DOF where necessary. We introduce a division scheme to ensure real-time haptic interaction performance. In the scheme, the deformation computation process is divided into two separate modules, which are then implemented on two separate threads. One module is dedicated to the update tasks of haptic renderings, which is implemented by extracting a sub-matrix from the pre-computed data matrix, and in this way haptic rendering process can be quickly refreshed at more than 1000 Hertz. The other module is dedicated to the task of deformation computation and visualization. The result of whole process is analyzed on an industrial example of design verification of a stamping mold.