1: \begin{abstract}

2: We compare the lattice results on QCD phase diagram for two and three

3: flavors with the hadron resonance gas model (HRGM) calculations. Lines of

4: constant energy density $\epsilon$ have been determined at different

5: baryo-chemical potentials $\mu_B$. For the strangeness chemical potentials

6: $\mu_S$, we use two models. In one model, we explicitly set $\mu_S=0$ for

7: all temperatures and baryo-chemical potentials. This assignment is used in

8: lattice calculations. In the other model, $\mu_S$ is calculated in

9: dependence on $T$ and $\mu_B$ according to the condition of vanishing

10: strangeness. We also derive an analytical expression for the dependence of

11: $T_c$ on $\mu_B/T$ by applying Taylor expansion of $\epsilon$. In both

12: cases, we compare HRGM results on $T_c-\mu_B$ diagram with the lattice

13: calculations. The agreement is 

14: excellent, especially when the trigonometric function of $\epsilon$ 

15: is truncated up to the same order as done in lattice simulations. For

16: studying the efficiency of the truncated Taylor expansion, we calculate the

17: radius of convergence. For zero- and second-order radii, the agreement with

18: lattice is convincing. Furthermore, we make predictions for QCD phase

19: diagram for non-truncated expressions and physical masses. These

20: predictions are to be confirmed by heavy-ion experiments and future lattice

21: calculations with very small lattice spacing and physical quark masses. 

22: \end{abstract}

23: