How should I calculate the test coverage for the manual testing I have done for my Software Project? Does it need some documentations like Test scenarios, test cases to calculate Test Coverage?
First of all you should answer the question: What would my test be covering? In white box testing tests usually cover the code lines. Since you do not have the code lines you should have something to be covered.
In black-box testing it might be the requirements. So that you have to have the requirements each of which would be tested by one or several tests. This approach would let you to calculate the coverage. Say 60% of requirements are covered.
Why do you need to answer that question? I would try to avoid that. Others are right by saying that usually for black box testing the answer would be requirement coverage. But what does requirement coverage mean? how meaningful a number is ? is 95% coverage consisting of 1 sunny-day test per requirement better or worst than 75% of multiple test at multiple levels per requirement ?
It's misleading the same as with code coverage numbers for white box testing- even if you have passed through all the branches and paths in the code you are still not safe if you haven't used your head in the process.
What can you do ? skip the percentages and present what is tested and how, shortly explain rather then giving one meaningless number. You can use visual heat maps that show areas of the product, percentage of coverage and quality of coverage.