1 | # rectangular_prism model |
---|
2 | # Note: model title and parameter table are inserted automatically |
---|
3 | r""" |
---|
4 | Definition |
---|
5 | ---------- |
---|
6 | |
---|
7 | This model provides the form factor, $P(q)$, for a hollow rectangular |
---|
8 | parallelepiped with a wall of thickness $\Delta$. The 1D scattering intensity |
---|
9 | for this model is calculated by forming the difference of the amplitudes of two |
---|
10 | massive parallelepipeds differing in their outermost dimensions in each |
---|
11 | direction by the same length increment $2\Delta$ (\ [#Nayuk2012]_ Nayuk, 2012). |
---|
12 | |
---|
13 | As in the case of the massive parallelepiped model (:ref:`rectangular-prism`), |
---|
14 | the scattering amplitude is computed for a particular orientation of the |
---|
15 | parallelepiped with respect to the scattering vector and then averaged over all |
---|
16 | possible orientations, giving |
---|
17 | |
---|
18 | .. math:: |
---|
19 | P(q) = \frac{1}{V^2} \frac{2}{\pi} \times \, \int_0^{\frac{\pi}{2}} \, |
---|
20 | \int_0^{\frac{\pi}{2}} A_{P\Delta}^2(q) \, \sin\theta \, d\theta \, d\phi |
---|
21 | |
---|
22 | where $\theta$ is the angle between the $z$ axis and the longest axis |
---|
23 | of the parallelepiped, $\phi$ is the angle between the scattering vector |
---|
24 | (lying in the $xy$ plane) and the $y$ axis, and |
---|
25 | |
---|
26 | .. math:: |
---|
27 | :nowrap: |
---|
28 | |
---|
29 | \begin{align*} |
---|
30 | A_{P\Delta}(q) & = A B C |
---|
31 | \left[\frac{\sin \bigl( q \frac{C}{2} \cos\theta \bigr)} |
---|
32 | {\left( q \frac{C}{2} \cos\theta \right)} \right] |
---|
33 | \left[\frac{\sin \bigl( q \frac{A}{2} \sin\theta \sin\phi \bigr)} |
---|
34 | {\left( q \frac{A}{2} \sin\theta \sin\phi \right)}\right] |
---|
35 | \left[\frac{\sin \bigl( q \frac{B}{2} \sin\theta \cos\phi \bigr)} |
---|
36 | {\left( q \frac{B}{2} \sin\theta \cos\phi \right)}\right] \\ |
---|
37 | & - 8 |
---|
38 | \left(\frac{A}{2}-\Delta\right) \left(\frac{B}{2}-\Delta\right) \left(\frac{C}{2}-\Delta\right) |
---|
39 | \left[ \frac{\sin \bigl[ q \bigl(\frac{C}{2}-\Delta\bigr) \cos\theta \bigr]} |
---|
40 | {q \bigl(\frac{C}{2}-\Delta\bigr) \cos\theta} \right] |
---|
41 | \left[ \frac{\sin \bigl[ q \bigl(\frac{A}{2}-\Delta\bigr) \sin\theta \sin\phi \bigr]} |
---|
42 | {q \bigl(\frac{A}{2}-\Delta\bigr) \sin\theta \sin\phi} \right] |
---|
43 | \left[ \frac{\sin \bigl[ q \bigl(\frac{B}{2}-\Delta\bigr) \sin\theta \cos\phi \bigr]} |
---|
44 | {q \bigl(\frac{B}{2}-\Delta\bigr) \sin\theta \cos\phi} \right] |
---|
45 | \end{align*} |
---|
46 | |
---|
47 | where $A$, $B$ and $C$ are the external sides of the parallelepiped fulfilling |
---|
48 | $A \le B \le C$, and the volume $V$ of the parallelepiped is |
---|
49 | |
---|
50 | .. math:: |
---|
51 | V = A B C \, - \, (A - 2\Delta) (B - 2\Delta) (C - 2\Delta) |
---|
52 | |
---|
53 | The 1D scattering intensity is then calculated as |
---|
54 | |
---|
55 | .. math:: |
---|
56 | I(q) = \text{scale} \times V \times (\rho_\text{p} - |
---|
57 | \rho_\text{solvent})^2 \times P(q) + \text{background} |
---|
58 | |
---|
59 | where $\rho_\text{p}$ is the scattering length density of the parallelepiped, |
---|
60 | $\rho_\text{solvent}$ is the scattering length density of the solvent, |
---|
61 | and (if the data are in absolute units) *scale* represents the volume fraction |
---|
62 | (which is unitless) of the rectangular shell of material (i.e. not including |
---|
63 | the volume of the solvent filled core). |
---|
64 | |
---|
65 | For 2d data the orientation of the particle is required, described using |
---|
66 | angles $\theta$, $\phi$ and $\Psi$ as in the diagrams below, for further details |
---|
67 | of the calculation and angular dispersions see :ref:`orientation` . |
---|
68 | The angle $\Psi$ is the rotational angle around the long *C* axis. For example, |
---|
69 | $\Psi = 0$ when the *B* axis is parallel to the *x*-axis of the detector. |
---|
70 | |
---|
71 | For 2d, constraints must be applied during fitting to ensure that the inequality |
---|
72 | $A < B < C$ is not violated, and hence the correct definition of angles is |
---|
73 | preserved. The calculation will not report an error if the inequality is *not* |
---|
74 | preserved, but the results may be not correct. |
---|
75 | |
---|
76 | .. figure:: img/parallelepiped_angle_definition.png |
---|
77 | |
---|
78 | Definition of the angles for oriented hollow rectangular prism. |
---|
79 | Note that rotation $\theta$, initially in the $xz$ plane, is carried out first, then |
---|
80 | rotation $\phi$ about the $z$ axis, finally rotation $\Psi$ is now around the axis of the prism. |
---|
81 | The neutron or X-ray beam is along the $z$ axis. |
---|
82 | |
---|
83 | .. figure:: img/parallelepiped_angle_projection.png |
---|
84 | |
---|
85 | Examples of the angles for oriented hollow rectangular prisms against the |
---|
86 | detector plane. |
---|
87 | |
---|
88 | |
---|
89 | Validation |
---|
90 | ---------- |
---|
91 | |
---|
92 | Validation of the code was conducted by qualitatively comparing the output |
---|
93 | of the 1D model to the curves shown in (Nayuk, 2012). |
---|
94 | |
---|
95 | |
---|
96 | References |
---|
97 | ---------- |
---|
98 | |
---|
99 | .. [#Nayuk2012] R Nayuk and K Huber, *Z. Phys. Chem.*, 226 (2012) 837-854 |
---|
100 | L. Onsager, Ann. New York Acad. Sci. 51, 627-659 (1949). |
---|
101 | |
---|
102 | |
---|
103 | Authorship and Verification |
---|
104 | ---------------------------- |
---|
105 | |
---|
106 | * **Author:** Miguel Gonzales **Date:** February 26, 2016 |
---|
107 | * **Last Modified by:** Paul Kienzle **Date:** December 14, 2017 |
---|
108 | * **Last Reviewed by:** Paul Butler **Date:** September 06, 2018 |
---|
109 | """ |
---|
110 | |
---|
111 | import numpy as np |
---|
112 | from numpy import pi, inf, sqrt |
---|
113 | |
---|
114 | name = "hollow_rectangular_prism" |
---|
115 | title = "Hollow rectangular parallelepiped with uniform scattering length density." |
---|
116 | description = """ |
---|
117 | I(q)= scale*V*(sld - sld_solvent)^2*P(q,theta,phi)+background |
---|
118 | P(q,theta,phi) = (2/pi/V^2) * double integral from 0 to pi/2 of ... |
---|
119 | (AP1-AP2)^2(q)*sin(theta)*dtheta*dphi |
---|
120 | AP1 = S(q*C*cos(theta)/2) * S(q*A*sin(theta)*sin(phi)/2) * S(q*B*sin(theta)*cos(phi)/2) |
---|
121 | AP2 = S(q*C'*cos(theta)) * S(q*A'*sin(theta)*sin(phi)) * S(q*B'*sin(theta)*cos(phi)) |
---|
122 | C' = (C/2-thickness) |
---|
123 | B' = (B/2-thickness) |
---|
124 | A' = (A/2-thickness) |
---|
125 | S(x) = sin(x)/x |
---|
126 | """ |
---|
127 | category = "shape:parallelepiped" |
---|
128 | |
---|
129 | # ["name", "units", default, [lower, upper], "type","description"], |
---|
130 | parameters = [["sld", "1e-6/Ang^2", 6.3, [-inf, inf], "sld", |
---|
131 | "Parallelepiped scattering length density"], |
---|
132 | ["sld_solvent", "1e-6/Ang^2", 1, [-inf, inf], "sld", |
---|
133 | "Solvent scattering length density"], |
---|
134 | ["length_a", "Ang", 35, [0, inf], "volume", |
---|
135 | "Shortest, external, size of the parallelepiped"], |
---|
136 | ["b2a_ratio", "Ang", 1, [0, inf], "volume", |
---|
137 | "Ratio sides b/a"], |
---|
138 | ["c2a_ratio", "Ang", 1, [0, inf], "volume", |
---|
139 | "Ratio sides c/a"], |
---|
140 | ["thickness", "Ang", 1, [0, inf], "volume", |
---|
141 | "Thickness of parallelepiped"], |
---|
142 | ["theta", "degrees", 0, [-360, 360], "orientation", |
---|
143 | "c axis to beam angle"], |
---|
144 | ["phi", "degrees", 0, [-360, 360], "orientation", |
---|
145 | "rotation about beam"], |
---|
146 | ["psi", "degrees", 0, [-360, 360], "orientation", |
---|
147 | "rotation about c axis"], |
---|
148 | ] |
---|
149 | |
---|
150 | source = ["lib/gauss76.c", "hollow_rectangular_prism.c"] |
---|
151 | have_Fq = True |
---|
152 | effective_radius_type = [ |
---|
153 | "equivalent cylinder excluded volume", "equivalent outer volume sphere", |
---|
154 | "half length_a", "half length_b", "half length_c", |
---|
155 | "equivalent outer circular cross-section", |
---|
156 | "half ab diagonal", "half diagonal", |
---|
157 | ] |
---|
158 | |
---|
159 | def random(): |
---|
160 | a, b, c = 10**np.random.uniform(1, 4.7, size=3) |
---|
161 | # Thickness is limited to 1/2 the smallest dimension |
---|
162 | # Use a distribution with a preference for thin shell or thin core |
---|
163 | # Avoid core,shell radii < 1 |
---|
164 | min_dim = 0.5*min(a, b, c) |
---|
165 | thickness = np.random.beta(0.5, 0.5)*(min_dim-2) + 1 |
---|
166 | #print(a, b, c, thickness, thickness/min_dim) |
---|
167 | pars = dict( |
---|
168 | length_a=a, |
---|
169 | b2a_ratio=b/a, |
---|
170 | c2a_ratio=c/a, |
---|
171 | thickness=thickness, |
---|
172 | ) |
---|
173 | return pars |
---|
174 | |
---|
175 | |
---|
176 | # parameters for demo |
---|
177 | demo = dict(scale=1, background=0, |
---|
178 | sld=6.3, sld_solvent=1.0, |
---|
179 | length_a=35, b2a_ratio=1, c2a_ratio=1, thickness=1, |
---|
180 | length_a_pd=0.1, length_a_pd_n=10, |
---|
181 | b2a_ratio_pd=0.1, b2a_ratio_pd_n=1, |
---|
182 | c2a_ratio_pd=0.1, c2a_ratio_pd_n=1) |
---|
183 | |
---|
184 | tests = [[{}, 0.2, 0.76687283098], |
---|
185 | [{}, [0.2], [0.76687283098]], |
---|
186 | ] |
---|