Repeated convolution of a continous uniform distribution

Thanks a lot to all who have contributed. I'm surprised that this
question sparked off so much discussion and interest :-)

The results are good enough for my purposes. Should I encounter
problems I will come back here.

Thanks again
weaker