Power Analysis for Mann Whitney U and/or variance ratio F tests

I'm interested in identifying the power of some tests I've run for my thesis. However, I'm not sure power analyses are possible and/or appropriate for Mann-Whitney U tests, and variance ratio F tests. Most of my samples are NOT normally distributed, although I custom generated new F distribution tables for the variance ration F tests, so maybe normality doesn't matter. If you can help me, I'd be so grateful. Can power tests be done for these analyses? How does one go about it? Thank you for your help.