A lot more simply:
$text = "èau? foo bar 88.";
var_dump(preg_replace('/p{M}/u', '',
Normalizer::normalize($text, Normalizer::FORM_D)));
gives:
string(16) "EauN foo bar 88."
The idea is to decompose the characters into the form where they're actually two Unicode code units, with an individual one for the accents. Then, you remove those accents.
p{M}
means "a character intended to be combined with another character (e.g. accents, umlauts, enclosing boxes, etc.)".
Note that this is not a definitive solution. It may be enough for the input text you have, but many characters, like ?
do not decompose, so this method will fail.
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…