Any aspirates that survived other changes became voiceless spirants in Ancient Quenya. Examples of this phonetic development are extremely numerous. Tolkien gave essentially identical descriptions of this sound change in both the Outline of Phonetic Development (OP1) from the 1930s and the Outline of Phonology (OP2) from the 1950s:

[OP1: PE19/30] The aspirated stops ph, th, kh remained probably in AQ; but in PQ (probably early) they were opened to strong voiceless spirants.

• (i) ph: produced a bilabial spirant [ꝑ].
• (ii) th: produced an interdental spirant [þ], more or less identical with the sound in English “thin”. This remained the learned and correct pronunciation in the classical period.
• (iii) kh: produced a back spirant [χ], as “ch” in Welsh. It was not markedly fronted by adjacent front vowels (e, i), at least no more than the corresponding stop.

[OP2: PE19/71] The aspirate stops ph, th, kh. These probably remained unchanged in AQ; but before the PQ period they had become strong voiceless spirants.

• (i) ph became a bilabial spirant [ꝑ]. ...
• (ii) th became the interdental spirant [þ], more or less identical with the sound of “th” in English “thin”.
• (iii) kh became the back spirant [χ], as “ch” in Welsh. It was not markedly fronted by adjacent front vowels ē̆, ī̆, at least no more than the corresponding stop k (or “ck” in English “thick”).

As noted above, the spirantalized ph first became the bilabial spirant ꝑ [ɸ], just as the voiced stop b initially spirantalized to ƀ [β]; later on these bilabial [ɸ], [β] became labio-dental [f], [v]. The spirantalized th, kh initially retained their place of articulation as þ, χ (IPA [θ] and [x]), but later these sounds (mostly) became s and h. Examples of all these spirantalizations of are trivial to find. There are some complexities in determining exactly when these sound changes took place, however.

To begin with, there was a similar sound change in Sindarin (and also in Noldorin of the 1930s). However, the aspirates underwent some Quenya-only changes before spirantalizing, notably aspirates pairs became voiceless stops; the opposite sound change occurred in Sindarin, where pairs of voiceless stops were aspirated (and later became voiceless spirants). Thus, the spirantalization of aspirates must have been a parallel sound change in the child languages rather than one belonging to Common Eldarin period.

In the notes above, Tolkien indicates this change occurred in late Ancient Quenya [AQ] or early Parmaquesta [PQ]. Whenever it happened, though, the spirantalization most likely occurred before velars were dentalized before [j], because initial khy > χy > (voiceless) hy was the usual phonetic development of this combination (OP1: PE19/34, OP2: PE19/75) and not khy > thy (> ?sy). The exacting timing of the dentalization of labialized-velars is itself unclear: it was after the voiced stops became spirants in the 1930s (since gy > ʒy > y, OP1: PE19/45) but before voiced stops spirantalized in the 1950s (since gy > dy > ly, OP2: PE19/75).

Since dentalization was an older sound change in the 1950s that it was in the 1930s, this implies the same is true of the spirantalization of aspirates. That’s consistent with what Tolkien said in the quotes above, where he stated that it was before the PQ period (probably late AQ) in the 1950s but occurred in early PQ in the 1930s. For simplicity, the Eldamo data model puts the spirantalization of aspirates in the Ancient Quenya period to account for various other sound changes that must have occurred after it.

Conceptual Development: Tolkien’s earliest conception of Primitive Elvish from the 1910s and 1920s had voiceless spirants rather than aspirates, so the spirantalization of the (non-existent) aspirates was not relevant.