1: \begin{abstract}

2: We systematically explore zero impact parameter collisions of white

3: dwarfs with the Eulerian adaptive grid code FLASH for 0.64+0.64

4: M$_{\odot}$ and 0.81+0.81 M$_{\odot}$ mass pairings. Our models span a

5: range of effective linear spatial resolutions from 5.2$\times$10$^{7}$

6: to 1.2$\times$10$^{7}$ cm.  However, even the highest resolution 

7: models do not quite achieve strict numerical convergence, due to the

8: challenge of properly resolving small-scale burning and energy

9: transport.  The lack of strict numerical convergence from these 

10: idealized configurations suggest that quantitative predictions of 

11: the ejected elemental abundances that are generated by binary white dwarf 

12: collision and merger simulations should be viewed with caution.

13: Nevertheless, the convergence trends do allow some patterns

14: to be discerned. We find that the 0.64+0.64 M$_{\odot}$ head-on collision model

15: produces 0.32 M$_{\odot}$ of \nickel[56] and 0.38 M$_{\odot}$ of

16: \silicon[28], while the 0.81+0.81 M$_{\odot}$ head-on collision model

17: produces 0.39 M$_{\odot}$ of \nickel[56] and 0.55 M$_{\odot}$ of

18: \silicon[28] at the highest spatial resolutions.  Both mass pairings

19: produce $\sim$0.2 M$_{\odot}$ of unburned \carbon[12]+\oxygen[16]. We

20: also find the 0.64+0.64 M$_{\odot}$ head-on collision begins 

21: carbon burning

22: in the central region of the stalled shock between the two white

23: dwarfs, while the more energetic \hbox{0.81+0.81 M$_{\odot}$} head-on

24: collision raises the initial post-shock temperature enough to burn the entire

25: stalled shock region to nuclear statistical equilibrium.

26: \end{abstract}

27: