Similarity experiments are conceived to study on existing tokamak facilities, characteristics of scenarios found on other devices or planned for new machines. The possibility of doing similarity experiments is linked to the physics processes studied and it gives in any case partial views which can be found in integrated way only on the planned devices. The paper presents scaling laws obtained to study pedestal physics, MHD limits and ELM behaviour, as well as bulk plasma confinement . The focus is on the dependence upon the aspect ratio and ion mass. The scaling laws are given in terms of plasma density(n), temperature(T), current(Ip), magnetic field(B) and input power (see definition in sec 1, Pinput) versus major radius(R), aspect ratio(A=R/a, R=major radius, a=minor radius) and ion mass(M)[12]. The introduction of ion mass is naturally included into the definition of the known set of dimensionless plasma physics parameters (q,ρ*T,ν*,βT) [6]. In a first instance, the scalings are obtained indipendent from energy confinement scaling laws. Further, if the scaling of heating power (Pheat) is obtained using the IPB(y,2) confinement scaling[6] or the DIIID/JET scaling [13-15] a sensible dependence of Pheat upon the geometry and aspect ratio is found. The scaling laws are obtained using the Kadomtsev-Lackner similarity scheme, where the alpha particle heating and atomic physics effects are neglected and the confinement is depending anly from the dimensionless set of parameters. The case of fusion reactor case is considered in detail in this paper: in this case the Kadomtsev-Lackner scheme is NOT valid since the alpha particles heating is relevant, leading to a new scaling parameter useful to characterize the fusion reactors at fixed Q.