Home / Math and Physics Files / Physics / Mechanics / new frames doc / Appendix H Euler Ang and Comp w
Euler unit vector exercise INSTALLED
DOCX · 22.8 KB
Open DOCX file
A short working note by Phil dated 3.14.17, installed in Appendix H of his mechanics frames document. It examines the unit vectors associated with the Euler angles φ, θ, ψ and asks whether they form bases for frames S and S'. He first tests the conjecture pψ = Lψ and finds it fails. He then expresses L in body-frame components using ω from (G.6.11) and concludes (L)z = pφ and (L)'z = pψ, as Goldstein claims.
AI-written summary; may contain errors. This description is approximate.
Extracted text (machine-read; may contain errors)
Euler unit vector exercise PhL 3.14.17
After looking at the Euler Angle unit vectors, I show (Question 3) that
(L)z = pφ
(L)'z = pψ
which is something Goldstein claims. This is installed now in Appendix H.
Based on (E.5.3), I think the following equations are valid
= -sinφ +cosφ
= -sinψ '+cosψ '
= -sinθ +cosθ
These unit vectors and equations do not appear anywhere in my Section G.5 (for some reason). Probably yes, but they are never really used as a basis by me.
Question 1: do these vectors form a basis for Frame S or for Frame S' ? Probably yes.
I could write them all out in terms of (,,) or (',',').
Question 2: Do , and form a basis for E3 ? Well, basically these are the tangent base vectors which I write there as en apart from scaling factors, and the en do form basis for x-space. Yes.
Question 3: In the top analysis, can I make these claims ?
Lφ = constant because φ is cyclic
Lψ = constant because ψ is cyclic
Using the above dot products I could write for example,
Lψ = L = -sinψ (L)'x + cosψ (L)'y
using body frame Frame S' evaluation. Is this related to Goldstein's
pψ = ∂L/∂ = ( cosθ + )I'3 ≡ aI'1 = (ω)'3I'3 ?
Recall that L = Iω so for sure,
(L)'i = (I)'ij(ω)'j = I'i(ω)'i
Then
(L)'x = (L)'1 = I'1(ω)'1
(L)'y = (L)'2 = I'2(ω)'2
Now recall that
(ω)'1 = sinθsinψ + cosψ
(ω)'2 = sinθcosψ - sinψ
(ω)'3 = cosθ + // Frame S' (G.6.11)
So I then know thaqt
(L)'x = (L)'1 = I'1(ω)'1 = I'1 [ sinθsinψ + cosψ]
(L)'y = (L)'2 = I'2(ω)'2 = I'2 [ sinθcosψ - sinψ ]
Then
Lψ = -sinψ (L)'x + cosψ (L)'y
= -sinψ [ sinθsinψ + cosψ] I'1 + cosψ [ sinθcosψ - sinψ ] I'2
But for the top we have I'2 = I'1 so write this as
Lψ = -sinψ [ sinθsinψ + cosψ] I'1 + cosψ [ sinθcosψ - sinψ ] I'1
= I'1 [ -sinθsin2ψ + sinθcos2ψ ] + I'1 [ -sinψ cosψ - cosψ sinψ ]
Compare this to
pψ = ∂L/∂ = ( cosθ + )I'3
I don't see any connection!!! So it looks like my conjecture that pψ = Lψ is no good.
Go back and review what has happened here. I wrote
Lψ = L = -sinψ (L)'x + cosψ (L)'y
= I'1 [ -sinθsin2ψ + sinθcos2ψ ] + I'1 [ -sinψ cosψ - cosψ sinψ ]
It seems more that, associated with angle ψ, we should have (L)'z being a constant of the motion, and not the quantity Lψ which really is orthogonal to the ' axis. You look for the axis about which your angle is a rotation. So now I think:
pψ = (L)'z ???
OK, write this out
(L)'z = (L)'3 = I'3 (ω')3 and yes this = pψ exactly.
Next conjecture that
(L)z is associated with φ, so should have (L)z = constant. Then
(L)z = L
Now how can I relate to the body frame basis vectors? I have failed to write this down anywhere in G.5. // But I have now added this and the result is
= (sinψsinθ)' + (cosψsinθ)' + (cosθ) '
Therefore
(L)z = L { (sinψsinθ)' + (cosψsinθ)' + (cosθ) ' }
= sinψsinθ (L)'1 + cosψsinθ (L)'2 + cosθ (L)'3
= sinψsinθ I'1(ω)'1 + cosψsinθ I'2(ω)'2 + cosθ I'3(ω)'3
= sinψsinθ I'1[ sinθsinψ + cosψ] + cosψsinθ I'2[ sinθcosψ - sinψ] + cosθ I'3[ cosθ + ]
= sinψsinθ I'1[ sinθsinψ + cosψ] + cosψsinθ I'1[ sinθcosψ - sinψ] + cosθ I'3[ cosθ + ]
= sinθ I'1 { sinθsin2ψ + cosψsinψ } + sinθ I'1 { sinθcos2ψ - sinψcosψ)
+ cosθ I'3[ cosθ + ]
= sinθ I'1 { sinθ } + cosθ I'3[ cosθ + ]
= sin2θI'1 + cos2θI'3 + cosθ I'3
= [sin2θI'1 + cos2θI'3 ] + cosθ I'3
Compare this with
pφ = (sin2θ I'1 + cos2θ I'3) + cosθ I'3
So I am now sure that
(L)z = pφ
(L)'z = pψ
Clarity keeps improving !