The letter ‘Æ’ evolved from the Classical Latin diphthong ‘AE’. With its long history, there are multiple models for its design. Nevertheless, there are some aspects worth highlighting.
The uppercase ‘Æ’
When merging ‘A’ and ‘E’, the type designer is faced with the challenge of determining which shape gets to play the “lead role”. Sacrificing the vertical stem allows the left ‘A’ part to stay more or less intact. This treatment, however, gives the capital ‘Æ’ a certain “vintage” vibe that can attract unwanted attention in a contemporary typeface. Unless the design is highly stylized, it is preferable to preserve the vertical stem of ‘E’.
A third option for capital ‘Æ’ is a compromise between the two — a slightly slanted middle stem. Though rarely seen, this variant has some merit — especially in text styles. It is an elegant solution to the many optical challenges, and does not attract unwanted attention if the slant is modest.
Balanced counterforms can be achieved by a slight lowering of the crossbar and widening of the apex in the ‘A’ part, as well as compressing the width of the ‘E’ part. The latter is a natural way to account for both upper and lower case variants needing a slight compression in the horizontal direction. The letter is, after all, one unit — not two.
The crossbars are not usually aligned in typographic typefaces, but may be rendered as a single uninterrupted stroke in handwriting. Additionally, seriffed designs can sometimes benefit from a top left serif to help fill the gap left by the strong diagonal.
The lowercase ‘æ’
Like the capital ‘Æ’, the minuscule ‘æ’ needs to be narrower than ‘a’ and ‘e’ combined, and the overall shape of ‘e’ should generally be preserved. The lowercase variant presents its own set of challenges: Branches must be thinned to avoid clogging. The lower part of the bowl usually requires some stretching to merge comfortably with the left side of ‘e’ and may necessitate a readjustment of its weight progression and contrast axis. I prefer to solve the ‘a’ part first. When it appears well balanced and similar enough to its origin, the ‘e’ is adjusted to match its width optically before final adjustments to the total width are made. The top of the ‘a’ bowl and the ‘e’ crossbar does not need to align, but in strongly chirographic type designs one may consider giving them a slight angle, so that they appear to meet in the middle.
Besides Norwegian, the ‘æ’ is used in Breton, Corsican, Danish, archaic English, Faroese, French, Genoese, Greenlandic, Icelandic, Ligurian, Lule Sami, Ripuarian and Southern Sami on the Norwegian side of the border. For those prone to supporting theoretical orthographies, one can also add Bornholmsk, Skånsk and Vallemål (where it may take an acute accent: Ǽ ǽ). Of dead languages, you can find it in Old Norse (including Ǽ ǽ for transliteration to Danish), Old English (including Northumbrian, where it may take a macron accent: Ǣ ǣ), and, additionally, as a ligature in Latin. It was also used in Ossetia for a short period between 1923 and 1937.
Typefaces that employ a single-story variant of ‘a’ — whether in the romans or in the cursives — runs the risk of confusing ‘æ’ and ‘œ’. Dynamic, writing-inspired, typefaces may find cues in Nordic and French handwriting models.
This is, however, not usually a viable solution in a geometric sans serif. Attempts to differentiate ‘æ’ and ‘œ’ with a clumsy middle stem has sadly become a common default solution for such designs.
Partially because French and its derived alphabets use both ‘œ’ and ‘æ’ (admittedly the ‘æ’ sees little use in day-to-day French), but also because the ‘œ’ appear foreign to us vikings, we recommend instead to render the lowercase ‘æ’ as a two-story variant, even if the ‘a’ only is of the single story kind. Some dissonance between ‘a’ and ‘æ’ is highly preferable to a malformed ‘æ’.