When I'm not sure what I'm doing on paper but I want neither to cross out previous equations if I turn out to be mistaken nor to be unable to plug them in later on - I use lots of sets of signs.

Most common are:
* Lower case English/latin (26 characters)
* Upper case English/latin (26 characters)
* Lower case Greek (24 characters)
The above have physical parameters assigned to them culturally like mass being 'm', velocity being 'v' and so on. So they are not very reusable for custom purposes unless you want to get confused.

Less common are:
* Upper case Greek (24 characters)
* Roman numerals (Very many characters)
I don't see them often, most of them are free to use without too much use in physics.

Very rare:
* Cursive/fancy latin. (26 characters)
Almost never used in maths and physics. Free to use mostly. Unless you can't make your cursive distinguishable which is a problem.

* Cursive upper Greek (24 characters)
* Cursive lower Greek (24 characters)
No idea why people don't use it more often. Any engineer should know it already, so it is 2 free sets for everyone with no effort.

* Chinese numerals
Quick to learn, feels related to roman numerals and I would use them if I wanted to do something similar to that which I had denoted with roman earlier. It is an infinite set.

* Hebrew abjad (22 characters)
Requires learning, but they are free 22 characters as you don't use alef everyday.

* Modified hiragana
* Modified katakana
You leave out signs that don't have all 5 variants of their row. (Keep まみむめも5-rower, but leave out やゆよbecause there is no ye and yi and there is no full 5 characters of this row!)
You are left with 8 sets of 5 characters and a few extras. Two sets actually because hiragana and katakana but pay attention to へ, leave its 5 characters out perhaps and you got at least 7x5.

You can add the tiny circle to any of them like in ぴ but do for all characters as needed. Even to those that legally can't have it. You doubled your set this way.
You can add tiny two lines to any of them び. Same as above, doubled the set.

Besides you have now 3 variants of the same root like ひ, び, ぴ for all 8 rows. ヒ, ビ and ピ are 3 extras of the same sound. You now can organize them either by dakuten, handakuten or none and by katakana or hiragana depending on your needs and how you assign relations in your equations. Extremely good. It's like a rectangle made of 1x2+1x2+1x2 or 2x3 whichever you like more, or a 2x2 square when you don't need more variables.

ひびぴ <--- look, how nice of a rectangle!
ヒビピ

* Chinese zodiac (120 variables)
Divided into yin and yang, one of the 12 animals and one of 5 elements.
2x12x5=120 <--- what a lovely cuboid!
You go with a standard order! Requires learning...

MORE SET IDEAS
* Slash your sign with a straight line - you doubled your set options. Like Ø or Ł but with an entire set.
* Give your characters an inverse of a 'hat'. Doubled your set options. Like Č or Š with an entire set.
* Encircle your characters to make a new set. Though it is tiring to look at.

NOT RECOMMENDED SETS
* Cyrillic
Shares too many characters with greek and latin.
If you implement it along with latin and greek you have 'e' and you have no idea whether it was from latin, greek, russian or just euler. Very bad choice!
* Chinese
Because unlike hiragana or katakana you cannot assign relations or a standard order (like alphabets) to things unless you use common radicals that could signify that these are related but that is it.
I would leave it as a miscellaneous set to use only if I can't find any relations between what I am calculating. A number trash bin where you throw numbers that you will not use later on or when very desperate.