Pulse palpation is an important procedure that allows a physician to rapidly assess the status of a patient's cardiovascular system. This paper explores the possibility of using vibrotactile stimuli to render fine temporal profiles of pulse pressure waves. A lightweight wearable vibrotactile glove, called Hap-pulse, is designed to render fine pulse waves through vibrotactile stimuli on users' fingertips. To preserve the fine features of original pulse waves, models are fitted from real pulse wave data (photoplethysmogram (PPG) pulse waveform database), using fourth-order polynomial functions. A square wave envelope mapping algorithm is proposed to produce vibration amplitudes of Linear Resonance Actuators (LRAs), which aims to render the detailed waveform of systolic and diastolic blood pressure states. Evaluation results suggest that Hap-pulse can render pulse waves with an average correlation coefficient 97.84%. To validate the distinguishability and fidelity of Hap-pulse's palpation rendering, a user study consisting of traditional Chinese medicine doctors and unskilled students is conducted. The correct recognition rate of identifying four typical pulse waves is 87.08% (doctors), 57.50% (untrained students) and 79.59% (trained students). These results indicate a novel application of rendering subtle pulse wave signals with vibrotactile gloves, which illustrates the potential of simulating patient palpation training in virtual or remote medical diagnosis.