1: \begin{abstract}

2: 

3: We have carried-out 98-level configuration-interaction / close-coupling (CI/CC) 

4: intermediate coupling frame transformation (ICFT) and Breit--Pauli $R$-matrix 

5: calculations for the electron-impact excitation of Be-like Al~$^{9+}$. The close

6: agreement that we find between the two sets of effective collision strengths

7: demonstrates the continued robustness of the ICFT method. 

8: On the other hand, a comparison of this data with previous 238-level 

9: CI/CC ICFT effective collision strengths shows that the results for

10: excitation up to $n=4$ levels are systematically and increasingly 

11: underestimated over a wide range of temperatures by $R$-matrix calculations 

12: whose close-coupling  expansion extends only to $n=4$ (98-levels). 

13: 

14: Thus, we find to be false a recent conjecture that the ICFT approach may 

15: not be completely robust. The conjecture was based upon a comparison of 

16: 98-level CI/CC Dirac $R$-matrix effective collision strengths for Al~$^{9+}$

17: with those from the 238-level CI/CC ICFT $R$-matrix calculations. The disagreement found

18: recently is due to a lack of convergence of the close-coupling expansion in the 

19: 98-level CI/CC Dirac work.

20: The earlier 238-level CI/CC ICFT work has a superior target

21: to  the 98-level CI/CC Dirac one and provides more accurate atomic data.

22: 

23: Similar considerations need to be made for other Be-like ions and for other sequences.

24: 

25: 

26: \end{abstract}

27: