We present lattice results for the vector and scalar form factors of the semileptonic decays D -> pi ell u_ell and D -> K ell u_ell in the physical range of values of squared four momentum transfer q^2, obtained with N_f=2 maximally twisted Wilson
fermions simulated at three different lattice spacings (a ~ 0.102 fm, 0.086 fm, 0.068 fm) with pion masses as light as 270 MeV and m_pi L gtrsim 4. The form factors are extracted using a double ratios strategy, which allows a good statistical accuracy and is independent of the vector current renormalization constant. The chiral/continuum extrapolation is performed through a simultaneous fit in the three variables (m_pi, q^2, a) using HMChPT formulae with additional O(a^2) terms that parametrically account for the lattice spacing dependence. Our results are in very good agreement with the experimental data in the full q^2 range for both D -> pi ell u_ell and D -> K ell u_ell. At zero momentum transfer we obtain f^{D->pi}(0) = 0.65(6)_{stat}(6)_{syst} and f^{D->K}(0) = 0.76(5)_{stat}(5)_{syst}, where the systematic error does not include the effects of quenching the strange and the charm quarks. Our findings are in good agreement with recent lattice calculations at N_f = 2+1.
We present lattice results for the form factors relevant in the K -> pion and D -> pion semileptonic decays, obtained from simulations with two flavors of dynamical twisted-mass fermions and pion masses as light as 260 MeV. For K -> pion decays we di
scuss the estimates of the main sources of systematic uncertainties, including the quenching of the strange quark, leading to our final result f+(0) = 0.9560 (57) (62). Combined with the latest experimental data, our value of f+(0) implies for the CKM matrix element |Vus| the value 0.2267 (5) (20) consistent with the first-row CKM unitarity. For D -> pion decays the application of Heavy Meson Chiral Perturbation Theory allows to extrapolate our results for both the scalar and the vector form factors at the physical point with quite good accuracy, obtaining a nice agreement with the experimental data. In particular at zero-momentum transfer we obtain f+(0) = 0.64 (5).
