The formation channels and predicted populations of double-white dwarfs (DWDs) are important because a subset will evolve to be gravitational-wave sources and/or progenitors of Type Ia supernovae. Given the observed population of short-period DWDs, we calculate the outcomes of common envelope evolution when convective effects are included. For each observed white dwarf in a DWD system, we identify all progenitor stars with an equivalent proto-WD core mass from a comprehensive suite of stellar evolution models. With the second observed white dwarf as the companion, we calculate the conditions under which convection can accommodate the energy released as the orbit decays, including (if necessary), how much the envelope must spin-up during the common envelope phase. The predicted post-CE final separations closely track the observed DWD orbital parameter space, further strengthening the view that convection is a key ingredient in common envelope evolution.

A widely used method for parameterizing the outcomes of common envelopes (CEs) involves defining an ejection efficiency, $\bar{\alpha}_{\mathrm{eff}}$, that represents the fraction of orbital energy used to unbind the envelope as the orbit decays. Given $\bar{\alpha}_{\mathrm{eff}}$, a prediction for the post-CE orbital separation is possible with knowledge of the energy required to unbind the primary's envelope from its core. Unfortunately, placing observational constraints on $\bar{\alpha}_{\mathrm{eff}}$ is challenging as it requires knowledge of the primary's structure at the onset of the common envelope phase. Numerical simulations have also had difficulties reproducing post-CE orbital configurations as they leave extended, but still bound, envelopes. Using detailed stellar interior profiles, we calculate $\bar{\alpha}_{\mathrm{eff}}$ values for a matrix of primary-companion mass pairs when the primary is at maximal extent in its evolution. We find that the ejection efficiency is most sensitive to the properties of the surface-contact convective region (SCCR). In this region, the convective transport timescales are often short compared to orbital decay timescales, thereby allowing the star to effectively radiate orbital energy and thus lower $\bar{\alpha}_{\mathrm{eff}}$. The inclusion of convection in numerical simulations of CEs may aid ejection without the need for additional energy sources as the orbit must shrink substantially further before the requisite energy can be tapped to drive ejection. Additionally, convection leads to predicted post-CE orbital periods of less than a day in many cases, an observational result that has been difficult to reproduce in population studies where $\bar{\alpha}_{\mathrm{eff}}$ is taken to be constant. Finally, we provide a simple method to calculate $\bar{\alpha}_{\mathrm{eff}}$ if the properties of the SCCR are known.