Data types store encoded information in very restrictive manners.
For example, integer values of the int data type are converted to binary integer numbers and thus, store only the discrete values allowed in the range of integers with that number of bits. On the other hand, the varchar data type keeps a character string as characters (without converting to binary) and thus takes more bits to represent the data and also allows more variety of input -- spaces, letters, special characters, etc. Of course, performing mathematical operations requires the int data type.
This means that 'A28' (I'm assuming that the 'A' is a letter and that this is not a hexadecimal number with 'A' as a digit) cannot be stored in a variable of the int data type.