The elastic–viscous–plastic method revisited