Many students are nervous to propose collaborations simply because of their academic station. I think awkward people may have more difficulty, but of course we have some advantages too.
Most scientists are pretty nerdy
There’s a certain personality type that is attracted to the academic life. Certainly non nerdy people can excel at research, but this career certainly appeals to nerds. Comfort yourself with the knowledge that the socially awkward population is over-represented in this area.
The least stressful way to propose a collaboration is via email. This has the benefit of being able to be proofread as well as read at the leisure of the target collaborator. In addition, you can attach links or articles if you so wish. It also gives the freedom of being whatever length you like.
Over-analyze
Don’t be afraid to over-analyze your proposal email or speech. It is critically important, and over analysis can help you improve your argument.
Suppress doubt
Remember: you both want the same thing. You both value the research and the knowledge. The only question is how to do the research, and who will be doing it.
Importance of formality
I have found that scientists don’t much care about formalities. After all, we spend our lives being specific and getting to the point. I am a firm believer that the content is more important than the delivery method. (See my post on politeness on science here.)
“But I’m only a grad student”
We all know imposter syndrome is a common issue in graduate school. Our status as students is not as important as we may imagine. Perhaps for an experimentalist collaborations may be more difficult due to resource limitations. But as computationalists there is no problem. All you need is your computer and a lit review and you’re good to go. Many experimentalists don’t have access to and modeler, and may not be aware that they need one. Modeling is a hot topic, so don’t be afraid to suggest it.
i am become computational 