Indexing a Welsh language document

I have a large document written in Welsh.  The Welsh language has a number of double letter consonants, 'ch', 'dd', 'ff', 'ng', 'll', 'ph', 'rh', and 'th'.  These are separate letters and should be indexed as such.  However in the words beginning with 'ch' gets added the the 'c' index whereas they should be in their own 'ch' index.

I thought I'd found the answer when I found a Swedish writer was having a similar problem with a Swedish document.  He was told that the 'index' field had a language code identifier.  Changing this to the code for Swedish solved his problem, apparently.

The language code for Welsh, I found, is 1106. Change the code in the field doesn't work for Welsh though.

Is there any way I can get my Index entries correctly?