q-bio0701006/sec2c.tex
1: \section{Comparison of shape, kinematic and stationary energy assumptions;
2:   validation}
3: \label{sect2c}
4: We now turn to a comparative study of the
5: force--displacement response of the circular arc and sinusoidal elasticas,
6: with the additional goal of gaining insight to matches between the
7: models and experimental data.
8: 
9: \subsection{The force-displacement response of
10:   circular-arc and sinusoidal elasticas subjected to different
11:   kinematic assumptions}     
12: \label{sect2c.1}
13: %------------- Figure -----
14: \begin{figure}[thb]
15: %
16: \psfrag{0l} [r][r]{\footnotesize  $0$}
17: \psfrag{1l} [r][r]{\footnotesize  $1$}
18: \psfrag{2l} [r][r]{\footnotesize  $2$}
19: \psfrag{3l} [r][r]{\footnotesize  $3$}
20: \psfrag{4l} [r][r]{\footnotesize  $4$} 
21: \psfrag{5l} [r][r]{\footnotesize  $5$} 
22: \psfrag{6l} [r][r]{\footnotesize  $6$} 
23: \psfrag{7l} [r][r]{\footnotesize  $7$} 
24: \psfrag{8l} [r][r]{\footnotesize  $8$} 
25: \psfrag{9l} [r][r]{\footnotesize  $9$} 
26: \psfrag{10l}[r][r]{\footnotesize $10$}
27: \psfrag{pi/2l} [r][r]{\tiny $\pi/2$}  
28: %
29: \psfrag{0}    [c][c]{\footnotesize $0$} 
30: \psfrag{pi/2} [c][c]{\footnotesize $\frac{\pi}{2}$} 
31: \psfrag{pi}   [c][c]{\footnotesize $\pi$} 
32: %
33: \psfrag{lmax}  [c][c]{\footnotesize 
34:                       $\bar\lambda_\mathrm{heel}=\theta_{0}/\sin\theta_{0}$} 
35: \psfrag{theta0}[c][c]{\footnotesize $\theta_0$} 
36: %
37: %
38: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
39: \psfrag{ 10l} [r][r]{\footnotesize $10$} 
40: \psfrag{ 20l} [r][r]{\footnotesize $20$} 
41: \psfrag{ 30l} [r][r]{\footnotesize $30$} 
42: \psfrag{ 40l} [r][r]{\footnotesize $40$} 
43: \psfrag{ 50l} [r][r]{\footnotesize $50$} 
44: %
45: \psfrag{ 1}    [c][c]{\footnotesize $1$} 
46: \psfrag{ 1.5}  [c][c]{\footnotesize $1.5$} 
47: \psfrag{ 2}    [c][c]{\footnotesize $2$} 
48: \psfrag{ 2.5}  [c][c]{\footnotesize $2.5$} 
49: \psfrag{ pi/2} [c][c]{\footnotesize $\frac{\pi}{2}$} 
50: %
51: \psfrag{f}    [c][c]{\footnotesize Tip force $f$} 
52: \psfrag{lmac} [c][c]{\footnotesize $\bar\lambda$} 
53: %
54: \psfrag{B1}    [l][l]{\scriptsize $B=1$} 
55: \psfrag{R1}    [l][l]{\scriptsize $R=1$} 
56: %
57: \psfrag{the030} [l][l]{\footnotesize
58:   $\theta_0{=}\displaystyle\frac{\pi}{6}$} 
59: \psfrag{the060} [l][l]{\footnotesize
60:   $\theta_0{=}\displaystyle\frac{\pi}{3}$}  
61: \psfrag{the090} [l][l]{\footnotesize
62:   $\theta_0{=}\displaystyle\frac{\pi}{2}$}  
63: \psfrag{the0120}[r][r]{\footnotesize
64:   $\theta_0{=}\displaystyle\frac{2\pi}{3}$}  
65: %
66: \psfrag{a} [l][l]{\normalsize $a$)} 
67: \psfrag{b} [l][l]{\normalsize $b$)} 
68: %
69: \centering
70: \includegraphics[width=14cm]{fig0.eps}
71: \caption{Circular-arc, inextensible elastica. $a$) Variation of the maximum 
72:   macro-stretch $\bar\lambda_\mathrm{heel}=\theta_{0}/\sin\theta_{0}$ with
73:   the initial angle $\theta_{0}$. $b$) Shift of the maximum value of
74:   the attainable 
75:   macro-stretch $\bar\lambda_\mathrm{heel}$ for selected values of the
76:   initial angle
77:   $\theta_{0}=\frac{\pi}{6},\frac{\pi}{3},\frac{\pi}{2},\frac{2\pi}{3}$,
78:   and for $B = 1$, $R = 1$.
79: }
80: \label{fig0}
81: \end{figure}
82: %------------- Figure -----
83: 
84: We first consider the force--stretch behavior of the circular-arc elastica
85: subject to the two additional kinematic assumptions and the stationary
86: strain energy assumption. To this end, we first relate the micro--tip
87: displacement, $g$, and the 
88: macro--stretch, $\bar\lambda$. The displacement between the ends of the
89: elastica is assumed to be dictated by macroscopic deformation in an
90: affine manner. That is, the macro--stretch $\bar\lambda$ is related to
91: the tip displacement $g$ by $\bar\lambda:= 1+  g/(2
92: R\sin\theta_{0})$ (see Figure \ref{figeg1}).
93: \footnote{For tissues with transverse isotropy, where the collagen fibrils
94:   (elasticas) are characterized by end-to-end vectors that are highly
95:   aligned, affinity of deformation is a good assumption. The
96:   alternative, fibril slippage, will be treated in a separate paper.}
97: In the studies to follow, the macro stretch will be used as the
98:   primary deformation variable controlling the force.
99: 
100: %------------- Figure -----
101: \begin{figure}[ht]
102: %
103: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
104: \psfrag{ 20l} [r][r]{\footnotesize $20$} 
105: \psfrag{ 40l} [r][r]{\footnotesize $40$} 
106: \psfrag{ 60l} [r][r]{\footnotesize $60$} 
107: \psfrag{ 80l} [r][r]{\footnotesize $80$} 
108: \psfrag{ 100l}[r][r]{\footnotesize $100$} 
109: %
110: \psfrag{ 1}   [c][c]{\footnotesize $1$} 
111: \psfrag{ 1.1} [c][c]{\footnotesize $1.1$} 
112: \psfrag{ 1.2} [c][c]{\footnotesize $1.2$} 
113: \psfrag{ 1.3} [c][c]{\footnotesize $1.3$} 
114: \psfrag{ 1.4} [c][c]{\footnotesize $1.4$} 
115: \psfrag{ 1.5} [c][c]{\footnotesize $1.5$} 
116: \psfrag{ 1.6} [c][c]{\footnotesize $1.6$} 
117: %
118: \psfrag{f} [c][c]{\footnotesize Tip force $f$} 
119: \psfrag{l} [l][l]{\footnotesize $\bar\lambda$} 
120: %
121: \psfrag{R0.4} [l][l]{\scriptsize $R=0.4$} 
122: \psfrag{R0.6} [l][l]{\scriptsize $R=0.6$} 
123: \psfrag{R0.8} [l][l]{\scriptsize $R=0.8$} 
124: \psfrag{R1}   [l][l]{\scriptsize $R=1$} 
125: %
126: \psfrag{B1}   [l][l]{\scriptsize $B=1$} 
127: \psfrag{B4}   [l][l]{\scriptsize $B=4$} 
128: \psfrag{B7}   [l][l]{\scriptsize $B=7$} 
129: \psfrag{B10}  [l][l]{\scriptsize $B=10$} 
130: %
131: \psfrag{a} [l][l]{\normalsize $a$)} 
132: \psfrag{b} [l][l]{\normalsize $b$)} 
133: %
134: \centering
135: \includegraphics[width=14cm]{fig1.eps}
136: \caption{Circular-arc. inextensible elastica. Sensitivity analysis of the
137:   $f-\bar\lambda$ curve to $a$) the variation of initial radius
138:   $R\in[0.4,1]$ and $b$) the variation of bending stiffness
139:   $B\in[1,10]$ for $\theta_{0}=\pi/2$. 
140: }
141: \label{fig1}
142: \end{figure}
143: %------------- Figure -----
144: 
145: For the circular-arc, inextensible elastica, according to
146: (\ref{elasticaforce1}), the tip force $f$ diverges
147: as the tip displacement $g$ approaches the value
148: $g_\mathrm{max}=2R(\theta_{0}-\sin\theta_{0})$. This implies that the 
149: maximum value of macro-stretch is
150: $\bar\lambda_\mathrm{heel}=\theta_{0}/\sin\theta_{0}\;,$ as shown also in
151: Figures \ref{fig0}a and \ref{fig0}b. The geometric parameter
152: $\theta_{0}$ thus has an unambiguous physical effect. The ordinates of
153: Figure \ref{fig0}a can be obtained as the locking stretch for each value of
154: $\theta_{0}$ as shown more transparently in Figure \ref{fig0}b for the 
155: $f{-}\bar\lambda$ response parametrized by $\theta_0$. For the circular-arc,
156: inextensible elastica, the shape of the $f{-}\bar\lambda$ curve
157: depends only on the initial radius $R$ and the bending stiffness
158: $B$. Figures \ref{fig1}a and \ref{fig1}b demonstrate that the
159: larger the initial radius $R$ or the smaller the bending stiffness
160: $B$, the sharper is the transition to divergence in $f$. Although
161: the sharpness of the transition can be tuned, the value of the
162: locking stretch remains
163: $\bar\lambda_\mathrm{heel}=\theta_0/\sin\theta_{0}$, and the
164: response beyond the heel region is asymptotic to a vertical line at
165: $\bar{\lambda}_\mathrm{heel}$. Of course, this divergent
166: $f{-}\bar\lambda$ response is non-physical, and considerably limits
167: the ability to match experiments on different collagenous materials possessing
168: distinct, and non-divergent, responses in the post-heel region.
169: 
170: 
171: %------------- Figure -----
172: \begin{figure}[ht]
173: %
174: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
175: \psfrag{ 20l} [r][r]{\footnotesize $20$} 
176: \psfrag{ 40l} [r][r]{\footnotesize $40$} 
177: \psfrag{ 60l} [r][r]{\footnotesize $60$} 
178: \psfrag{ 80l} [r][r]{\footnotesize $80$} 
179: \psfrag{ 100l}[r][r]{\footnotesize $100$} 
180: %
181: \psfrag{ 1l}  [r][r]{\footnotesize  $1$} 
182: \psfrag{ 1.1l}  [r][r]{\footnotesize  $1.1$} 
183: \psfrag{ 1.2l}  [r][r]{\footnotesize  $1.2$} 
184: %
185: \psfrag{ 1}   [c][c]{\footnotesize $1$} 
186: \psfrag{ 1.1} [c][c]{\footnotesize $1.1$} 
187: \psfrag{ 1.2} [c][c]{\footnotesize $1.2$} 
188: \psfrag{ 1.3} [c][c]{\footnotesize $1.3$} 
189: \psfrag{ 1.4} [c][c]{\footnotesize $1.4$} 
190: \psfrag{ 1.5} [c][c]{\footnotesize $1.5$} 
191: \psfrag{ 1.6} [c][c]{\footnotesize $1.6$} 
192: \psfrag{ 1.8} [c][c]{\footnotesize $1.8$} 
193: \psfrag{ 2} [c][c]{\footnotesize $2$} 
194: \psfrag{ 3} [c][c]{\footnotesize $3$} 
195: \psfrag{ 4} [c][c]{\footnotesize $4$} 
196: \psfrag{ 5} [c][c]{\footnotesize $5$} 
197: %
198: \psfrag{f} [c][c]{\footnotesize Tip force $f$} 
199: \psfrag{l} [l][l]{\footnotesize $\bar\lambda$} 
200: \psfrag{lam} [c][c]{\footnotesize $\lambda=r\theta/(R\theta_{0})$ } 
201: %
202: \psfrag{EA1}   [l][l]{\scriptsize $EA=1$} 
203: \psfrag{EA34}  [l][l]{\scriptsize $EA=34$} 
204: \psfrag{EA67}  [l][l]{\scriptsize $EA=67$} 
205: \psfrag{EA100} [l][l]{\scriptsize $EA=100$} 
206: %
207: \psfrag{B1}   [l][l]{\scriptsize $B=1$} 
208: \psfrag{R1}   [l][l]{\scriptsize $R=1$} 
209: \psfrag{the0} [l][l]{\scriptsize $\theta_{0}=\pi/3$} 
210: %
211: \psfrag{the030} [r][r]{\footnotesize $\theta_0{=}\frac{\pi}{6}$} 
212: \psfrag{the060} [l][l]{\footnotesize $\theta_0{=}\frac{\pi}{3}$}  
213: \psfrag{the077} [l][l]{\footnotesize $\theta_{0_\mathrm{cr}}{\approx}1.342$}  
214: %\psfrag{app}    [l][l]{\tiny         $\approx77^{\circ}$}  
215: \psfrag{app}    [l][l]{\tiny }  
216: \psfrag{the090} [l][l]{\footnotesize $\theta_0{=}\frac{\pi}{2}$}  
217: 
218: %
219: \psfrag{a} [l][l]{\normalsize $a$)} 
220: \psfrag{b} [l][l]{\normalsize $b$)} 
221: %
222: \centering
223: \includegraphics[width=14cm]{fig2.eps}
224: \caption{Circular-arc elastica surrounded by a planar incompressible
225:   medium. $a$) Dependence of the micro--stretch $\lambda$
226:   on the macro--stretch $\bar\lambda$ and the initial angle $\theta_{0}$. 
227:   $b$) Sensitivity analysis of the $f{-}\bar\lambda$ curve to
228:   the variation of the axial stiffness  $EA\in[1,100]\;.$
229:   curve.
230: }
231: \label{fig2}
232: \end{figure}
233: %------------- Figure -----
234: 
235: The variation of the micro-stretch, $\lambda$, with macro-stretch,
236: $\bar\lambda$, for a 
237: circular-arc elastica embedded in a planar incompressible medium  is
238: depicted in Figure \ref{fig2}a. We 
239: draw attention to the compression of the
240: elastica for a regime of 
241: deformation characterized by small values of $\bar{\lambda}$, and
242: discussed in Remark 4. Following the approach outlined there we have
243: solved for $\theta_0 = \theta_{0_\mathrm{cr}}$ such that for all $\theta_0 <
244: \theta_{0_\mathrm{cr}}$ we have
245: $\mathrm{d}\lambda/\mathrm{d}\bar{\lambda} > 0$. 
246: Explicitly we have
247: $\lambda=\theta(\bar\lambda)r(\bar\lambda)/(R\theta_0)$ with  
248: $r(\bar\lambda)=R(1-\cos(\theta_{0}))/(2\bar\lambda) +
249: \bar\lambda^{3}R\sin^{2}(\theta_{0})/(2(1-\cos(\theta_{0}))$ and 
250: $\theta(\bar\lambda)=\sin^{-1}(\bar\lambda
251: R\sin\theta_{0}/r(\bar\lambda))$ . Setting the derivative
252: $\mathrm{d}\lambda/\mathrm{d}\bar\lambda|_{\bar\lambda=1}=0$ and
253: solving the resulting equation for $\theta_{0}$, we obtained the maximum: $\theta_{0}\le  \theta_{0_\mathrm{cr}} \approx 1.342$. As
254: clearly shown in Figure 
255: \ref{fig2}a, for the values of the initial angle
256: $\theta_{0}\le \theta_{0_\mathrm{cr}}$ the compressive regime of
257: micro-stretch is avoided. For this reason our subsequent
258: investigations are restricted to 
259: $\theta_{0}\le \theta_{0_\mathrm{cr}}$ for the circular-arc elastica
260: embedded in a planar incompressible medium.    
261: 
262: The sensitivity of the tip force, $f$, to the axial stiffness,
263: $EA$, for $B = 1$,  $R = 1$ and $\theta_{0}=\pi/3$ is depicted in
264: Figure \ref{fig2}b for the circular-arc elastica surrounded by a planar
265: incompressible medium. A decrease in the axial 
266: stiffness translates, as expected, to a decrease of the slope. Clearly, the axial
267: stiffness dominates the slope of the $f{-}\bar{\lambda}$ curve. 
268: In contrast to the inextensible case we observe neither a long
269: toe region nor a distinguishable heel region. Larger values of
270: $\bar\lambda$ are attainable. As the axial 
271: stiffness is increased, however, the  $f{-}\bar{\lambda}$ curve enters
272: the high (but still increasing) slope regime about $\bar{\lambda}_\mathrm{heel}
273: =\theta_0/\sin\theta_{0}$ without exhibiting much of a toe region. This, of
274: course, limits the use of this model.    
275: 
276: %------------- Figure -----
277: \begin{figure}[ht]
278: %
279: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
280: \psfrag{ 20l} [r][r]{\footnotesize $20$} 
281: \psfrag{ 40l} [r][r]{\footnotesize $40$} 
282: \psfrag{ 60l} [r][r]{\footnotesize $60$} 
283: \psfrag{ 80l} [r][r]{\footnotesize $80$} 
284: \psfrag{ 100l}[r][r]{\footnotesize $100$} 
285: %
286: \psfrag{ 1l}  [r][r]{\footnotesize  $1$} 
287: \psfrag{ 1.1l}  [r][r]{\footnotesize  $1.1$} 
288: \psfrag{ 1.2l}  [r][r]{\footnotesize  $1.2$} 
289: %
290: \psfrag{ 1}   [c][c]{\footnotesize $1$} 
291: \psfrag{ 1.1} [c][c]{\footnotesize $1.1$} 
292: \psfrag{ 1.2} [c][c]{\footnotesize $1.2$} 
293: \psfrag{ 1.3} [c][c]{\footnotesize $1.3$} 
294: \psfrag{ 1.4} [c][c]{\footnotesize $1.4$} 
295: \psfrag{ 1.5} [c][c]{\footnotesize $1.5$} 
296: \psfrag{ 1.6} [c][c]{\footnotesize $1.6$} 
297: \psfrag{ 1.8} [c][c]{\footnotesize $1.8$} 
298: \psfrag{ 2} [c][c]{\footnotesize $2$} 
299: \psfrag{ 3} [c][c]{\footnotesize $3$} 
300: \psfrag{ 4} [c][c]{\footnotesize $4$} 
301: \psfrag{ 5} [c][c]{\footnotesize $5$} 
302: %
303: \psfrag{f} [c][c]{\footnotesize Tip force $f$} 
304: \psfrag{l} [l][l]{\footnotesize $\bar\lambda$} 
305: \psfrag{lam} [c][c]{\footnotesize $\lambda=2r\theta/(\pi R)$ } 
306: %
307: \psfrag{EA1}   [l][l]{\scriptsize $EA=1$} 
308: \psfrag{EA34}  [l][l]{\scriptsize $EA=34$} 
309: \psfrag{EA67}  [l][l]{\scriptsize $EA=67$} 
310: \psfrag{EA100} [l][l]{\scriptsize $EA=100$} 
311: %
312: \psfrag{B1}   [l][l]{\scriptsize $B=1$} 
313: \psfrag{R1}   [l][l]{\scriptsize $R=1$} 
314: \psfrag{the0} [l][l]{\scriptsize $\theta_{0}=\frac{\pi}{2}$} 
315: %
316: \psfrag{a} [l][l]{\normalsize $a$)} 
317: \psfrag{b} [l][l]{\normalsize $b$)} 
318: %
319: \centering
320: \includegraphics[width=7cm]{fig3.eps}
321: \caption{Circular-arc elastica with stationary energy. Sensitivity analysis
322:   of the $f{-}\bar\lambda$ curve to the variation of the axial stiffness
323:   $EA\in[1,100]\;.$   
324: }
325: \label{fig3}
326: \end{figure}
327: %------------- Figure -----
328: 
329: The material parameters used for the circular-arc elastica
330: with stationarity of strain energy were the same as for the planar
331: incompressible medium case. As in the planar incompressible case a decrease in
332: axial stiffness has a strong, depressing influence on slope of the
333: $f{-}\bar{\lambda}$ response. The elastica can be extended to
334: $\bar{\lambda}>\theta_{0}/\sin\theta_{0}\;.$ 
335: With an increase in axial stiffness, the $f{-}\bar\lambda$ curve approaches
336: the behavior of the inextensible circular-arc elastica (see Figure
337: \ref{fig3}). The circular-arc elastica attaining a
338: stationary strain energy possesses a number of favorable properties:
339: The toe region exists and its slope can be tuned
340: by the bending stiffness. The location of the heel region is uniquely
341: determined by the initial angle $\theta_{0}$ as $\bar{\lambda}_\mathrm{heel}
342: =\theta_0/\sin\theta_{0}$. The slope of the
343: post-heel region can be adjusted by the axial stiffness $EA$ as shown in
344: Figure \ref{fig3}. The stationary strain energy assumption with clearly
345: identifiable parameters thus serves as a promising model to match with
346: experimental data.        
347: 
348: 
349: In Figure \ref{fig4} we compare all three cases of the circular-arc
350: elastica. Distinct values $EA=34,67,100$ are assigned to the axial
351: modulus of the planar incompressible and stationary energy elasticas. The
352: $f{-}\bar\lambda$ curve of the stationary energy case approaches the
353: inextensible one by ``rotating'' about the heel just below
354: $\bar{\lambda}_\mathrm{heel}$. However, this occurs with no discernible 
355: difference in the curves for $\lambda$ values smaller than the
356: heel. In case of the elastica surrounded by a planar incompressible medium,
357: however, the stiffening in the $f{-}\bar\lambda$ behavior is
358: different. Owing to larger values of micro--stretch in the initial
359: stages, the location of the heel shifts to smaller values of
360: $\bar{\lambda}$.
361: 
362: %------------- Figure -----
363: \begin{figure}[ht]
364: %
365: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
366: \psfrag{ 20l} [r][r]{\footnotesize $20$} 
367: \psfrag{ 40l} [r][r]{\footnotesize $40$} 
368: \psfrag{ 60l} [r][r]{\footnotesize $60$} 
369: \psfrag{ 80l} [r][r]{\footnotesize $80$} 
370: \psfrag{ 100l}[r][r]{\footnotesize $100$} 
371: %
372: \psfrag{ 1l}  [r][r]{\footnotesize  $1$} 
373: \psfrag{ 1.1l}  [r][r]{\footnotesize  $1.1$} 
374: \psfrag{ 1.2l}  [r][r]{\footnotesize  $1.2$} 
375: %
376: \psfrag{ 1}   [c][c]{\footnotesize $1$} 
377: \psfrag{ 1.1} [c][c]{\footnotesize $1.1$} 
378: \psfrag{ 1.2} [c][c]{\footnotesize $1.2$} 
379: \psfrag{ 1.3} [c][c]{\footnotesize $1.3$} 
380: \psfrag{ 1.4} [c][c]{\footnotesize $1.4$} 
381: \psfrag{ 1.5} [c][c]{\footnotesize $1.5$} 
382: \psfrag{ 1.6} [c][c]{\footnotesize $1.6$} 
383: \psfrag{ 1.8} [c][c]{\footnotesize $1.8$} 
384: \psfrag{ 2} [c][c]{\footnotesize $2$} 
385: \psfrag{ 3} [c][c]{\footnotesize $3$} 
386: \psfrag{ 4} [c][c]{\footnotesize $4$} 
387: \psfrag{ 5} [c][c]{\footnotesize $5$} 
388: %
389: \psfrag{f} [c][c]{\footnotesize Tip force $f$} 
390: \psfrag{l} [l][l]{\footnotesize $\bar\lambda$} 
391: \psfrag{lam} [c][c]{\footnotesize $\lambda=2r\theta/(\pi R)$ } 
392: %
393: \psfrag{EA}   [l][l]{\scriptsize $EA{=}34,67,100$} 
394: %
395: \psfrag{ie} [r][r]{\tiny Inext.} 
396: \psfrag{me} [r][r]{\tiny Stat. Energy } 
397: \psfrag{ic} [r][r]{\tiny Incomp.} 
398: %
399: \psfrag{B1}   [l][l]{\scriptsize $B{=}1$} 
400: \psfrag{R1}   [l][l]{\scriptsize $R{=}1,\:\theta_{0}{=}\pi/3$} 
401: %
402: \centering
403: \includegraphics[width=7cm]{fig4.eps}
404: \caption{Comparison of circular-arc elasticas subjected to different
405:   constraints. In the planar incompressible and stationary strain energy cases,
406:   three different values are assigned to the axial stiffness
407:   $EA=34,67,100$, corresponding to increasingly stiff
408:   $f{-}\bar{\lambda}$ response.
409: }
410: \label{fig4}
411: \end{figure}
412: %------------- Figure -----
413: \noindent
414: In the foregoing parameter study, we solely considered circular-arc
415: elasticas with two kinematic assumptions and the stationary strain
416: energy assumption. In what
417: follows, we present an analogous parameter sensitivity study for 
418: the sinusoidal geometry. In contrast to the circular-arc elastica, the
419: reference shape of a sinusoidal elastica is governed by two
420: parameters: the amplitude $a_0$ and the half--wave 
421: length $l_0$ (see Figure \ref{sin1}). The ratio $a_0/l_0$, however, cannot be
422: arbitrarily chosen. According to the results reported by
423: \citet{Daleetal:72}, this ratio is limited to 
424: values smaller than $0.1$. Accounting for this fact in 
425: the studies to follow the ratio has been chosen as
426: $a_0/l_0< 0.2\;,$ which will allow us to consider values slightly
427: larger than the experimental observations. The macro--stretch, $\bar\lambda$,
428: remains the primary deformation measure, and is now related
429: to the tip displacement, $g$, by $\bar\lambda = 1+  g/l_0$.   
430: 
431: 
432: First, we consider a sinusoidal elastica with the additional
433: global inextensibility assumption given in
434: (\ref{gloinextensibility}). In Figure \ref{fig5}a the influence of the 
435: ratio $a_0/l_0 \in [0.05,0.2]$  on the $f{-}\bar\lambda$ curve is
436: depicted while keeping the material parameters fixed at $B=1$ and $EA=1$. This
437: ratio proves crucial in determining the value of 
438: stretch at which the heel occurs. The higher the ratio $a_0/l_0$, the
439: longer the toe region preceding the heel. In other words, this
440: parameter determines the value of $\bar{\lambda}$ where the influence
441: of the bending mechanism starts to 
442: diminish and the axial extension begins to govern the
443: $f{-}\bar\lambda$ curve. In
444: order to demonstrate the sensitivity of the $f{-}\bar\lambda$ curve to the
445: bending stiffness, the ratio of bending stiffness to axial stiffness,
446: $B/EA$, is varied from $1$ to $4$ (Figure \ref{fig5}b). An increase
447: in the ratio $B/EA$ scales the curve's ordinates ($f$-values), and
448: therefore the 
449: transition in the heel region becomes more gradual. However, the value
450: of the locking stretch is not influenced by the changes in the ratio $B/EA$.
451: %------------- Figure -----
452: \begin{figure}[ht]
453: %
454: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
455: \psfrag{ 20l} [r][r]{\footnotesize $20$} 
456: \psfrag{ 40l} [r][r]{\footnotesize $40$} 
457: \psfrag{ 60l} [r][r]{\footnotesize $60$} 
458: \psfrag{ 80l} [r][r]{\footnotesize $80$} 
459: \psfrag{ 100l}[r][r]{\footnotesize $100$} 
460: %
461: \psfrag{ 1}    [c][c]{\footnotesize $1$} 
462: \psfrag{ 1.01} [c][c]{\footnotesize $1.01$} 
463: \psfrag{ 1.02} [c][c]{\footnotesize $1.02$} 
464: \psfrag{ 1.03} [c][c]{\footnotesize $1.03$} 
465: \psfrag{ 1.04} [c][c]{\footnotesize $1.04$} 
466: \psfrag{ 1.06} [c][c]{\footnotesize $1.06$} 
467: \psfrag{ 1.08} [c][c]{\footnotesize $1.08$} 
468: \psfrag{ 1.1}  [c][c]{\footnotesize $1.1$} 
469: %
470: \psfrag{f} [c][c]{\footnotesize Tip force $f$} 
471: \psfrag{l} [l][l]{\footnotesize $\bar\lambda$} 
472: %
473: \psfrag{a/l0.05} [l][l]{\scriptsize $a_0/l_0=0.05$} 
474: \psfrag{a/l0.1}  [l][l]{\scriptsize $a_0/l_0=0.1$} 
475: \psfrag{a/l0.15} [l][l]{\scriptsize $a_0/l_0=0.15$} 
476: \psfrag{a/l0.2}  [l][l]{\scriptsize $a_0/l_0=0.2$} 
477: %
478: \psfrag{B}   [l][l]{\scriptsize $B=1$} 
479: \psfrag{EA}  [l][l]{\scriptsize $EA=1$} 
480: %
481: \psfrag{B/EA1}   [l][l]{\scriptsize $B/EA=1$} 
482: \psfrag{B/EA2}   [l][l]{\scriptsize $B/EA=2$} 
483: \psfrag{B/EA3}   [l][l]{\scriptsize $B/EA=3$} 
484: \psfrag{B/EA4}   [l][l]{\scriptsize $B/EA=4$} 
485: %
486: \psfrag{a} [l][l]{\normalsize $a$)} 
487: \psfrag{b} [l][l]{\normalsize $b$)} 
488: %
489: \centering
490: \includegraphics[width=14cm]{fig5.eps}
491: \caption{Sinusoidal inextensible elastica.  Comparison of the
492:   $f{-}\bar\lambda$ curves for globally inextensible sinusoidal
493:   elasticas having different $a$) $a_0/l_0$ and $b$) $B/EA$ ratios.  
494: }
495: \label{fig5}
496: \end{figure}
497: %------------- Figure -----
498: 
499: In the last two cases we consider the planar incompressible and stationary energy
500: sinusoidal elasticas. Figures \ref{fig6}a and \ref{fig7}a
501: present the influence of the change in ratio $a_0/l_0$ on 
502: the $f{-}\bar\lambda$ curves of the respective cases. Like the 
503: inextensible case the ratio $a_0/l_0$ is varied within
504: the interval $[0.05,0.2]$ while the value of the ratio
505: $EA/B$ is kept fixed at $30$. Clearly, the
506: $f{-}\bar\lambda$ curves for the 
507: planar incompressible and stationary energy cases do not exhibit a sharp
508: transition to
509: stiffening behavior. This is in contrast with the inextensible
510: case in Figure \ref{fig5}. Variation of the ratio $a_0/l_0$ does not
511: cause significant change in the shape of the curves. 
512: 
513: %------------- Figure -----
514: \begin{figure}[ht]
515: %
516: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
517: \psfrag{ 100l}[r][r]{\footnotesize $100$} 
518: \psfrag{ 200l}[r][r]{\footnotesize $200$} 
519: \psfrag{ 300l}[r][r]{\footnotesize $300$} 
520: \psfrag{ 400l}[r][r]{\footnotesize $400$} 
521: \psfrag{ 500l}[r][r]{\footnotesize $500$} 
522: %
523: \psfrag{ 1}    [c][c]{\footnotesize $1$} 
524: \psfrag{ 1.5}  [c][c]{\footnotesize $1.5$} 
525: \psfrag{ 2}    [c][c]{\footnotesize $2$} 
526: \psfrag{ 2.5}  [c][c]{\footnotesize $2.5$} 
527: %
528: \psfrag{f} [c][c]{\footnotesize Tip force $f$} 
529: \psfrag{l} [l][l]{\footnotesize $\bar\lambda$} 
530: %
531: \psfrag{a/l0.02} [l][l]{\scriptsize $a_0/l_0=0.02$} 
532: \psfrag{a/l0.08} [l][l]{\scriptsize $a_0/l_0=0.08$} 
533: \psfrag{a/l0.14} [l][l]{\scriptsize $a_0/l_0=0.14$} 
534: \psfrag{a/l0.2}  [l][l]{\scriptsize $a_0/l_0=0.2$} 
535: \psfrag{a/l0.1}  [l][l]{\scriptsize $a_0/l_0=0.1$} 
536: %
537: \psfrag{EA/B30}  [l][l]{\scriptsize $EA/B=30$} 
538: \psfrag{EA/B120}  [l][l]{\scriptsize $EA/B=120$} 
539: \psfrag{EA/B210}  [l][l]{\scriptsize $EA/B=210$} 
540: \psfrag{EA/B300}  [l][l]{\scriptsize $EA/B=300$} 
541: %
542: \psfrag{a} [l][l]{\normalsize $a$)} 
543: \psfrag{b} [l][l]{\normalsize $b$)} 
544: %
545: \centering
546: \includegraphics[width=14cm]{fig6.eps}
547: \caption{Sinusoidal elastica surrounded by a planar incompressible
548:  medium. Comparison of the $f{-}\bar\lambda$ curves for  different
549:  $a$) $a_0/l_0$ and $b$) $EA/B$ ratios.
550: }
551: \label{fig6}
552: \end{figure}
553: %------------- Figure -----
554: The sensitivity of the $f{-}\bar\lambda$ curves for the separate cases
555: to changes in material parameters $EA$ and $B$ is presented in Figures 
556: \ref{fig6}b and \ref{fig7}b, respectively. The ratio
557: $EA/B$ varies in the range $[30,300]$. The axial
558: stiffening is clearly reflected in the curves. No striking
559: shape change is observed. We draw attention to the fact that the 
560: $f{-}\bar\lambda$ curves in Figures \ref{fig6} and \ref{fig7} for the
561: planar incompressible and stationary energy cases of the sinusoidal elastica 
562: are quite similar. The reasons for this similarity have
563: been already outlined in Remark 8.
564: 
565: %------------- Figure -----
566: \begin{figure}[ht]
567: %
568: \psfrag{ 0l}  [r][r]{\footnotesize  $0$} 
569: \psfrag{ 100l}[r][r]{\footnotesize $100$} 
570: \psfrag{ 200l}[r][r]{\footnotesize $200$} 
571: \psfrag{ 300l}[r][r]{\footnotesize $300$} 
572: \psfrag{ 400l}[r][r]{\footnotesize $400$} 
573: \psfrag{ 500l}[r][r]{\footnotesize $500$} 
574: %
575: \psfrag{ 1}    [c][c]{\footnotesize $1$} 
576: \psfrag{ 1.5}  [c][c]{\footnotesize $1.5$} 
577: \psfrag{ 2}    [c][c]{\footnotesize $2$} 
578: \psfrag{ 2.5}  [c][c]{\footnotesize $2.5$} 
579: %
580: \psfrag{f} [c][c]{\footnotesize Tip force $f$} 
581: \psfrag{l} [l][l]{\footnotesize $\bar\lambda$} 
582: %
583: \psfrag{a/l0.02} [l][l]{\scriptsize $a_0/l_0=0.02$} 
584: \psfrag{a/l0.08} [l][l]{\scriptsize $a_0/l_0=0.08$} 
585: \psfrag{a/l0.14} [l][l]{\scriptsize $a_0/l_0=0.14$} 
586: \psfrag{a/l0.2}  [l][l]{\scriptsize $a_0/l_0=0.2$} 
587: \psfrag{a/l0.1}  [l][l]{\scriptsize $a_0/l_0=0.1$} 
588: %
589: \psfrag{EA/B30}  [l][l]{\scriptsize $EA/B=30$} 
590: \psfrag{EA/B120}  [l][l]{\scriptsize $EA/B=120$} 
591: \psfrag{EA/B210}  [l][l]{\scriptsize $EA/B=210$} 
592: \psfrag{EA/B300}  [l][l]{\scriptsize $EA/B=300$} 
593: %
594: \psfrag{a} [l][l]{\normalsize $a$)} 
595: \psfrag{b} [l][l]{\normalsize $b$)} 
596: %
597: \centering
598: \includegraphics[width=14cm]{fig7.eps}
599: \caption{Sinusoidal elastica deforming by attaining a stationary strain
600:  energy state. Comparison of the $f{-}\bar\lambda$ curves for 
601:  different $a$) $a_0/l_0$ and $b$) $EA/B$ ratios.  
602: }
603: \label{fig7}
604: \end{figure}
605: %------------- Figure -----
606: 
607: 
608: \subsection{Comparison with experiment}
609: In the preceding section the sensitivities of the $f-\bar{\lambda}$
610: curves to geometric and material parameters have been discussed for
611: both the circular-arc and the sinusoidal elasticas subjected to two
612: additional kinematic assumptions, and the stationary strain energy
613: assumption. In this section we carry out a comparison with 
614: data reported by \citet{freed+doehring05} (Figure \ref{fig8}). 
615: %------------- Figure -----
616: \begin{figure}[thb]
617: %
618: \psfrag{ 0l}    [r][r]{\footnotesize  $0$} 
619: \psfrag{ 0.2l}  [r][r]{\footnotesize  $0.2$} 
620: \psfrag{ 0.4l}  [r][r]{\footnotesize  $0.4$} 
621: \psfrag{ 0.6l}  [r][r]{\footnotesize  $0.6$} 
622: \psfrag{ 0.8l}  [r][r]{\footnotesize  $0.8$} 
623: \psfrag{ 1l}    [r][r]{\footnotesize  $1$} 
624: \psfrag{ 1.2l}  [r][r]{\footnotesize  $1.2$} 
625: \psfrag{ 1.4l}  [r][r]{\footnotesize  $1.4$} 
626: %
627: \psfrag{ 1}    [c][c]{\footnotesize $1$} 
628: \psfrag{ 1.1}  [c][c]{\footnotesize $1.1$} 
629: \psfrag{ 1.2}  [c][c]{\footnotesize $1.2$} 
630: \psfrag{ 1.4}  [c][c]{\footnotesize $1.4$} 
631: \psfrag{ 1.6}  [c][c]{\footnotesize $1.6$} 
632: \psfrag{ 1.8}  [c][c]{\footnotesize $1.8$} 
633: \psfrag{ 2}    [c][c]{\footnotesize $2$} 
634: %
635: \psfrag{f} [c][c]{\footnotesize Nominal Stress $\:[MPa]$} 
636: \psfrag{l} [l][l]{\footnotesize $\bar\lambda\:[-]$} 
637: %
638: \psfrag{exp} [l][l]{\scriptsize Experiment}  
639: \psfrag{sin_ie} [l][l]{\scriptsize Sinusoidal}  
640: \psfrag{cir_me} [l][l]{\scriptsize Circular-Arc}  
641: %
642: \centering
643: \includegraphics[width=7cm]{fig8.eps}
644: \caption{Simulations of experimental data by the inextensible
645:   sinusoidal elastica ($a_0/l_0=0.245\;, B/EA=25\:\mathrm{mm}^{-1}$) and the
646:   circular-arc elastica attaining a stationary strain energy state
647:   ($R=0.013\:\mathrm{mm}\;,\theta_{0}=4\pi/ 15\;,
648:   EA/B=7\times 10^{6}\:\mathrm{mm}^{-2}$). In the elastica models $R,
649:   a_0$ and $l_0$ were in mm
650:   $B$ in $\mathrm{N}/\mathrm{mm}$ and $EA$ in $\mathrm{MPa}/\mathrm{mm}$.  
651: }
652: \label{fig8}
653: \end{figure}
654: %------------- Figure -----
655: These data correspond to uniaxial extension experiments on five
656: chordae tendineae from procine mitral valves. They demonstrate a long
657: toe region relative to the maximum stretch in each experiment. At
658: the heel $(\bar\lambda_\mathrm{heel}\approx 1.13)$,  the nominal
659: stress--stretch curve stiffens sharply to a larger
660: slope. From the results in Figures \ref{fig0}--\ref{fig7} of the
661: preceding parameter study, it is apparent that this behavior can only
662: be captured either by the inextensible sinusoidal elastica, or the
663: circular-arc elastica attaining a stationary energy state. Figure
664: \ref{fig8} compares the experiment with these two models with the
665: material parameters given in the caption. Both the sinusoidal and the
666: circular-arc models successfully match the data in the toe region. The
667: inextensible sinusoidal model can also predict the upturning region,
668: but its stiffness beyond the heel region rapidly diverges and fails to
669: match the experimental results. In the case of the circular-arc model,
670: the value of 
671: $\theta_{0}$ can be analytically determined from the
672: macro-stretch value at the heel. We solve for $\theta_0$ such that
673: $\theta_0/\sin\theta_{0} = \bar\lambda_\mathrm{heel}=1.13$.
674: This gives the initial angle $\theta_{0}\approx 4\pi/ 15$, and
675: the ratio of the axial and bending stiffness $EA/B$ can be tuned to
676: match the slopes of the regions just preceding and succeeding
677: the heel region. The initial radius $R$ is varied to match the
678: sharpness of the slope change at the heel region. The 
679: comparison of the stationary strain energy circular-arc elastica and the
680: experimental data clearly illustrates that the proposed model
681: quantitatively captures the experimental data with just a few
682: parameters: $\theta_0, R, B$ and $EA$, all of which are very
683: well-motivated physically. Clearly, other such experimental data can be matched
684: without difficulty.
685: 
686: 
687: