Abstract: To study the influence of heating parameters, including heating pulse duration, heat flux density and heating energy, on surface temperature signals, a 2D model for carbon fiber reinforced plastic sample and a 3D model for aluminum sample were investigated and solved by finite element method(FEM). Calculation results showed that the differential temperature between defect and sound areas depended on heating energy, and it got greater while the heating energy increasing. Maximum temperature contrast, differential temperature peak time and best observation time were less sensitive to the change of heat flux density, but affected by heating pulse duration. Maximum temperature contrast was less affected by heating parameters when materials were carbon fiber reinforced plastics(CFRP), so it had more relations to defects and could be used for quantitative analysis of defects. Great heat flux density and short heating pulse was a better choice for detecting defects in metal materials with greater thermal diffusivity, such as aluminum.