A comprehensive database of detonation run-up distances (chi(DDT)) in unobstructed tubes/channels is compiled. Eight fuels are included, i.e., hydrogen, methane, ethane, ethylene, acetylene, propane, butane and carbon monoxide. Oxygen is used as oxidizer with different diluents under a wide range of tube/channel sizes. In total 559 points are collected and analyzed. The global chi(DDT) trends observed in the data as a function of the initial conditions (i.e., fuel type, equivalence ratio (phi), hydraulic diameter (D-h), and pressure (p)) reveal that chi(DDT) DDT decreases with increasing p, decreasing Dh, and phi -> 1. The correlation of the normalized run-up distance chi(DDT)/D-h with geometrical parameters and fundamental combustion properties shows that most of the variance in the experimental data can be captured by the ratio of D-h to the flame thickness (D-h/delta(f)), the expansion ratio (sigma - 1) and that of the Chapman-Jouguet to the laminar flame speed (V-CJ/sigma S-L). With these, a non-linear (NLM) and a logarithmic model (LM) are proposed using a non-linear least squares regression to fit a user-defined function. The NLM and LM are shown to outperform the widely used Silvestrini and Dorofeev models by a large margin since they are capable of explaining similar to 70-80% of the variance in the experimental chi(DDT)/D-h data in contrast to only similar to 20% of the original Silvestrini and Dorofeev models. The poor performance of the original models is due to the limited amount of data used to determine the models constants. An update to these constants is also carried out using the database collected in this work resulting in a notable improvement of their predictive capabilities over the original models. The database is publicly available so that it can be used freely to guide future research in the combustion community (e.g., by identifying conditions where there is a lack of data), and as a test bed for further data-driven model development.