1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3:
4: \begin{document}
5: \title{Confinement of supernova
6: explosions in a collapsing cloud}
7: \author{M. M. Kasliwal, R.V.E. Lovelace, and J.R. Houck}
8: \affil{Department of Astronomy,
9: Cornell University, Ithaca, NY 14853; MMK32@cornell.edu;
10: RVL1@cornell.edu; JRH13@cornell.edu}
11:
12: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
13: \begin{abstract}
14:
15: We analyze the
16: confining effect of cloud
17: collapse on an expanding
18: supernova shockfront.
19: We solve the differential
20: equation for the forces on the
21: shockfront due to ram pressure,
22: supernova energy, and gravity.
23: We find that the expansion of
24: the shockfront is slowed and in fact reversed
25: by the collapsing cloud.
26: Including radiative losses and a potential
27: time lag between supernova explosion and cloud
28: collapse shows that the expansion is reversed at
29: smaller distances as compared to the non-radiative
30: case.
31: We also consider the case of multiple
32: supernova explosions at the center
33: of a collapsing cloud.
34: For instance, if we scale our
35: self-similar solution to a single supernova of energy
36: $10^{51}$ ergs occurring when a cloud of initial
37: density $10^{2}$ H/cm$^{3}$ has collapsed by $50\%$,
38: we find that the shockfront is confined to $\sim 15$ pc in
39: $\sim 1$ Myrs.
40: Our calculations are pertinent to
41: the observed unusually
42: compact non-thermal radio emission
43: in blue compact dwarf galaxies (BCDs). More generally, we
44: demonstrate the potential of a collapsing cloud to confine
45: supernovae, thereby explaining how dwarf galaxies would exist
46: beyond their first generation of star formation.
47:
48: \end{abstract}
49:
50: \keywords{galaxies: individual (SBS 0335-052, Henize 2-10) --
51: galaxies: dwarf -- supernova remnant -- galaxies: starburst -- stars:
52: dwarf nova}
53:
54:
55: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
56:
57: \section{Introduction}
58:
59: Blue Compact Dwarf galaxies (BCDs) are
60: observed to be experiencing
61: intense star formation in a spatially compact
62: region. The timescale for this star formation
63: episode of mass $\sim$ 10$^{8} M_{\sun}$ is
64: only a few million years.
65: Massive stars have short lifetimes and thus
66: supernova (SN) explosions may occur concurrently in
67: regions of cloud collapse. It is quite probable that this
68: occurs in blue compact dwarf galaxies. In this paper,
69: we study the confining effect of cloud collapse on
70: an expanding supernova shockfront.
71:
72: The forces acting on the shockfront
73: include those due to supernova energy,
74: ram pressure and gravity.
75: When a supernova shockfront expands into a stationary cloud
76: (Sedov 1946, Taylor 1950, Bisnovatyi-Kogan \& Silich 1995),
77: the accreted mass does not contribute to a change in the
78: momentum flux. In a region of cloud collapse, not only
79: is mass accreted at a faster rate, it also contributes
80: to a change in the momentum flux. This additional ram pressure
81: leads to the confinement of a supernova shockfront in a collapsing
82: cloud. The potential of ram pressure to confine stellar winds of
83: a noncentral OB star to generate a steady-flow situation has been shown
84: (Dopita 1981). We demonstrate that ram pressure can also confine expanding
85: supernova shockfronts with pressures as high as $\sim$ 10$^{-9}$ dynes/cm$^{2}$
86: as compared to stellar wind pressure which is $\sim$ 10$^{-12}$ dynes/cm$^{2}$
87: (Dopita et al 1981). Our self-similar solution describes the
88: time evolution of a central supernova explosion in a collapsing cloud.
89:
90: We start with a discussion of the
91: gravitational collapse of a cloud,
92: and then obtain the equation for
93: the expansion of a SN shock in
94: the cloud. We first neglect
95: radiative losses and then
96: include them. Next, we consider
97: a time lag between when the supernova explodes
98: and when the cloud begins to collapse.
99: Finally, we consider the
100: case of multiple SN explosions
101: occurring at the center
102: of a two-component collapsing cloud.
103: The inner component has higher
104: density and collapses rapidly
105: to produce the central stellar
106: population. A few million years
107: later, the resulting supernova
108: shockfronts collide with the
109: infalling, less dense,
110: outer component of the cloud.
111:
112:
113: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
114: \section{Model}
115:
116:
117: We first consider a simple model
118: for the free-fall gravitational collapse
119: of a cloud with negligible pressure.
120: The gravitational force on a
121: differential spherical shell with
122: initial radius $R_0$ and radius $R(t)$, is
123: \begin{equation}
124: \label{CC1}
125: dM(R)~\frac{d^2R}{dt^2}=
126: -\frac{GM(R_0)~dM(R)}{R^2}~,
127: \end{equation}
128: where G is the universal gravitational
129: constant and $M(R_0)$ is the initial
130: mass enclosed by this shell.
131: $
132: M(R_0)=4\pi R_0^3\rho_0/3,
133: $
134: where $\rho_0$ is the initial
135: density of the cloud which is
136: assumed to be uniform.
137: Solving this equation,
138: assuming
139: the initial velocity is zero,
140: gives
141: \begin{equation}
142: \label{VcR}
143: {dR \over dt}=
144: -\left[2GM(R_0)\left(\frac{1}{R}-
145: \frac{1}{R_0}\right)\right]^{1/2}~.
146: \end{equation}
147: The free-fall time of the cloud is
148: \begin{eqnarray}
149: \label{tff}
150: {t_{ff}}&=&\int_{0}^{R_0} \frac{dR}{-dR/dt}
151: =\left(\frac {3\pi}{32G\rho_0}\right)^{1/2}
152: \nonumber\\
153: &\approx& {5.2\times 10^6 {\rm yr}}~{(n_{2})^{-1/2}}~,
154: \end{eqnarray}
155: where $n_{2}$ is density in units of
156: 10$^2$ H/cm$^{3}$.
157: The free-fall time is
158: independent of the initial radius. Thus,
159: the collapse is homologous
160: as shown in Figure 1.
161:
162: The velocity of the shell hitting the shock
163: front contributes to the ram pressure that
164: slows down the front.
165: We normalize the time
166: and radius as $\tau={t}/{t_{ff}}$
167: and $\overline{R}={R}/{R_0}$.
168: Hence we obtain
169: \begin{equation}
170: \label{tau}
171: \tau=\frac{2}{\pi}
172: \int_{\overline{R}}^{1} dr^\prime
173: \left(\frac{r^\prime}{1-r^\prime}\right)^{1/2}~.
174: \end{equation}
175: This integral gives
176: a cumbersome expression for
177: $\overline{R}(\tau)$.
178: Therefore, we
179: approximate the dependence
180: as
181: \begin{equation}
182: \label{ftau}
183: \overline{R} \approx (1-\tau^2)^{1/2}\equiv f(\tau)~.
184: \end{equation}
185: Thus the velocity of the cloud at the
186: shockfront radius, $R_{sh}$, is
187: \begin{equation}
188: \label{Vc}
189: {V_{cl}(\tau)=
190: -\frac{\pi R_{sh}[1-f(\tau)]^{1/2}}
191: {2t_{ff}{[f(\tau)]^{3/2}}}}~.
192: \end{equation}
193:
194: Accretion of mass just outside
195: the shock front acts to
196: increase in density.
197: Mass conservation gives
198: \begin{equation}
199: \label{delM}
200: \Delta M =
201: 4\pi R_0^2 \Delta R_0\rho_0 =
202: 4\pi R^2 \Delta R\rho(t) ~.
203: \end{equation}
204: Using $R=R_{0}f(\tau)$ and $\Delta R=
205: \Delta R_{0}f(\tau)$ gives
206: \begin{equation}
207: \label{rho}
208: \rho(\tau)=\frac{\rho_0}{[f(\tau)]^3}~.
209: \end{equation}
210:
211: There are three forces experienced by the shock front:
212: (i) The energy released by the supernova
213: explosion $E_{SN}$ inside the shell gives
214: an outward directed pressure
215: $p=(\gamma - 1)E_{SN}/(\frac{4}{3}\pi R_{sh}^3)$
216: or force $4\pi R_{sh}^2 p$. We use $\gamma = 5/3$.
217: (ii) The ram pressure of the cloud shell
218: hitting the shock front results in
219: an inward directed force (or pressure).
220: When the direction of motion of the
221: external medium and the
222: shockfront are the same, the ram pressure acts only if
223: the speed of the medium is greater than that
224: of the shock front.
225: This is accounted for by a
226: Heaviside function, $H$, of the difference between
227: the shockfront velocity and cloud velocity
228: (with $H(>0)=1$ and $H(<0)=0$).
229: (iii) The gravitational force is of course inward.
230:
231: The ambient pressure of the interstellar medium
232: is $\sim$ 10$^{-10}$ dynes/cm$^{2}$
233: for a density $\sim$ 100 ${\rm H/cm}^{3}$ and temperature $\sim$ 10$^{4}$ K.
234: For a typical speed $\sim$ 100 km/s, the ram pressure is $\sim$
235: 10$^{-8}$ dynes/cm$^{2}$. For this reason, we neglect the ambient
236: pressure of the interstellar medium.
237:
238: The rate at which the SN shockfront accretes mass is proportional
239: to the relative velocity between the cloud and the shockfront. Thus,
240:
241: \begin{equation}
242: \label{dMtau}
243: \frac{dM_{sh}}{dt} = 4\pi{R_{sh}}^2~\rho~U~,
244: \end{equation}
245: where
246: \begin{equation}
247: U \equiv {{dR_{sh}}\over{dt}}-V_{cl}~.
248: \end{equation}
249:
250: The accreted mass initially moved
251: at the cloud velocity. It thus
252: contributes to a net change in the momentum flux.
253: For the case where
254: the SN explosion and cloud collapse
255: begin at the same time we have
256: \begin{equation}
257: \label{DEQ}
258: \frac{d}{dt}\left(M_{sh}\frac{dR_{sh}}{dt}\right) =
259: 4\pi R_{sh}^2\left[\frac{E_{SN}}{2 \pi R_{sh}^3}
260: + \rho H(U)~U~V_{cl} \right]
261: -\frac{GM_{sh}^2}{2R_{sh}^2}~.
262: \end{equation}
263:
264: We normalize time by the free-fall time
265: and the shockfront radius by the
266: Sedov-Taylor radius at the free-fall
267: time (within a factor of 1.24); that
268: is,
269: $\overline{R}_{sh}={R_{sh}}/{R_{ST}(t_{ff})}$, where
270: \begin{equation}
271: R_{ST}(t_{ff})\equiv ((\gamma - 1)E_{SN}t_{ff}^2/\rho_0)^{1/5}
272: \approx 52~{\rm pc}~
273: (E_{51})^{1/5}
274: (n_{2})^{-2/5}~.
275: \end{equation}
276: We normalize mass by
277: $\overline{M}_{sh}={M_{sh}}
278: /[4 \pi\rho_0 R_{ST}^3(t_{ff})/3]$.
279:
280:
281: We define a `momentum' variable
282: \begin{equation}
283: \label{P}
284: {\cal P} \equiv \overline{M}_{sh}
285: \frac{d\overline{R}_{sh}}{d\tau}~.
286: \end{equation}
287:
288:
289: Rewriting equation (\ref{dMtau}) and equation (\ref{DEQ})
290: in terms of the normalized parameters gives
291:
292: \begin{equation}
293: \frac{d\overline{M}_{sh}}{d\tau} =
294: \frac{3{\overline{R}_{sh}}^{2}}{f^3}\left(\frac{d\overline{R}_{sh}}{d\tau}+\frac{\pi\overline{R}_{sh}({1-f})^{1/2}}{2f^{3/2}}\right)~.
295: \end{equation}
296:
297:
298:
299: \begin{equation}
300: \frac{d{\cal P}}{d\tau} =
301: \frac{{9}}{4\pi{\overline{R}_{sh}}}
302: -\frac{3\pi{\overline{R}_{sh}}^{3}~
303: ({1-f})^{1/2}~H(\overline{U})~
304: \overline{U}}{2~f^{9/2}}
305: -\frac{\pi^2\overline{M}_{sh}^{2}}{16{\overline{R}_{sh}}^2}~,
306: \end{equation}
307: where
308: \begin{equation}
309: \label{Ubar}
310: \overline{U}=
311: \frac{\cal P}{\overline{M}_{sh}}
312: +\frac{{\pi\overline{R}_{sh}}
313: ({1-f})^{1/2}}{2{f}^{3/2}}~.
314: \end{equation}
315: Thus, we have a system of three
316: non-linear coupled first order equations.
317: For the initial conditions
318: we assume at an early time, $R_{sh}$
319: and ${dR_{sh}}/{d\tau}$
320: are given by the standard
321: Sedov-Taylor solution for a stationary
322: cloud. We use the initial condition
323: that at $\tau=10^{-6}$,
324: $\overline{R}_{sh}=0.0049$,
325: $\overline{M}_{sh}=1.2 \times {10}^{-7}$
326: and $\cal{P}$$ = 0.00024$.
327: Figure 2 shows $\overline{R}_{sh}(\tau)$
328: and the corresponding Sedov-Taylor solution.
329: The infalling cloud at first slows
330: the shock expansion and later reverses
331: its motion.
332: %%
333:
334: \section{Influence of Radiative Losses}
335:
336: In the above analysis we assumed
337: that the energy within the
338: SN shock was constant, but
339: this is valid only for relatively
340: short times.
341: We next take into account the
342: radiative losses using the
343: pressure driven snowplow
344: model of Cioffi,
345: McKee, and Bertschinger (1988).
346: We find that the supernova
347: is confined to a much smaller
348: region in a shorter time.
349: We use the mean pressure
350: $\overline{p}_{rl}$ including
351: radiative losses derived by
352: Cioffi et al. (1988)
353: which can be written as
354: \begin{equation}
355: \label{Prad}
356: \overline{p}_{rl}=
357: \frac{\alpha~E_{SN}^{5/3}}{R_{sh}^5~t^{4/9}~
358: \rho^{10/9}~\eta_M^{4/7}}~.
359: \end{equation}
360: where magnitude of $\alpha$ is $2.3 \times 10^{-17}$ and $\eta_M$ is
361: metallicity in units of solar metallicity.
362: Thus $\overline{p}_{rl}$ replaces
363: our earlier pressure $p=E_{SN}/(2\pi R_{sh}^3)$.
364: Next, we assume that the
365: metallicity is solar metallicity
366: and we use equation (\ref{rho})
367: for density as a function of time.
368: The differential equation including
369: the radiative losses is
370: \begin{equation}
371: \label{DEQrad}
372: \frac{d}{dt}\left(M_{sh}\frac{dR_{sh}}{dt}\right) =
373: 4\pi~R_{sh}^2\left(\overline{p}_{rl}
374: - \rho H(U)~U~V_{cl} \right)
375: -\frac{GM_{sh}^2}{2R_{sh}^2}~.
376: \end{equation}
377: We continue to normalize time
378: by the free-fall time.
379: However, we now normalize the radius by
380: \begin{eqnarray}
381: \label{Rrad}
382: {R_{rl}}&=&\left({\frac{3\alpha~E_{SN}^{5/3}~
383: t_{ff}^{14/9}}{\rho_{0}^{19/9}}}\right)^{1/7}
384: \nonumber\\
385: &\approx& {11.8 {\rm pc}~(E_{51})^{5/21}~(n_{2})^{-26/63}}~.
386: \end{eqnarray}
387:
388: We again obtain a self-similar
389: differential equation and we
390: solve it using the initial
391: condition that the shockfront follows
392: the Sedov solution at $\tau=10^{-6}$.
393: We use this initial condition
394: because the
395: transition from the Sedov solution to
396: the pressure driven snowplow radiative
397: shockfront occurs at
398: \begin{equation}
399: t_{PDS}\approx 1.33 \times 10^4~{\rm yr}~
400: \frac{E_{51}^{3/14}}{n_0^{4/7} },
401: \end{equation}
402: (Cioffi et al. 1988).
403: For $E=10^{51}$ ergs and $n_0=10^2 {\rm H/cm}^{3}$,
404: $t_{PDS} \approx 1700 $ yr $ \sim t_{ff}/3000$.
405:
406:
407: The SN explosion may occur at
408: a time $\Delta t$ after the cloud collapse has begun.
409: The velocity and density of the shells
410: hitting the shock front is then larger.
411: We take the zero of time $t$
412: or $\tau=t/t_{ff}$ to be the
413: time when the supernova explodes so that
414: only equation (\ref{ftau}) needs to be
415: modified and replaced by
416: $f(\tau)=\overline{R}=
417: [1-(\tau+\Delta\tau)^2]^{1/2}
418: $, where $\Delta \tau =\Delta t/t_{ff}$ is
419: termed the ``lag.''
420:
421: Figure 3 shows sample results for different lags which shows that the
422: shock expansion is more strongly decelerated the longer the lag.
423: For no time lag, the shockfront is confined to $\overline{R}_{sh}$
424: = 1.99 and $\tau$ = 0.51. In comparison, for 50\% time lag,
425: the shockfront is confined to $\overline{R}_{sh}$ = 1.33 and
426: $\tau$ = 0.23 and for 80\% time lag, the shockfront is
427: confined to $\overline{R}_{sh}$ = 0.75 and $\tau$ = 0.09.
428:
429:
430: \section{Multiple Supernova Explosions}
431:
432: We now consider a more complete physical
433: picture. Suppose that there is a
434: localized clump of density $10^4$ H/cm$^3$
435: in a cloud of density $10^2$ H/cm$^3$.
436: Suppose that the ambient outer cloud
437: and the inner clump begin to free-fall.
438: The clump collapses in $\sim 0.5$ Myrs and forms
439: supernovae in another $2-3.5$ Myrs (lifetime of $30-24
440: M_{\sun}$ star). The free-fall time
441: of the cloud is $\sim 5$ Myrs and thus, the time lag
442: between when the supernova explodes and when the cloud
443: begins to collapse ranges from $0.5 t_{ff}$ to
444: $0.8 t_{ff}$.
445:
446: When the localized clump collapses, the initial mass
447: function is such that many low mass (lm) stars
448: form at the same time as a few high mass
449: stars.
450: The distribution is dN/dM $\propto$ M$^{-2.35}$ (Salpeter 1955).
451: If we take
452: the high mass (hm) threshold at $> 20 M_{\sun}$,
453: then $M_{lm} \sim
454: 200N_{hm}M_{\sun}$.
455: The mass of the low mass stars contributes
456: to the gravitational force in addition to the forces described
457: in equation(\ref{DEQrad}). The right-hand side of this
458: equation becomes
459: \begin{equation}
460: 4\pi R_{sh}^2\left[\overline{p}_{rl}
461: - \rho H(U)~U~V_{cl} \right]
462: -\frac{GM_{sh}^2}{2R_{sh}^2}
463: -\frac{GM_{sh}M_{lm}}{R_{sh}^2}~.
464: \end{equation}
465: We find however that the
466: gravitational force contribution
467: of the low mass stars does not significantly
468: alter the trajectory of the SN shockfront.
469:
470: \section{Discussion}
471:
472: Regarding the stability of the shockfront,
473: notice that the
474: Rayleigh-Taylor
475: instability occurs in a
476: stratified region
477: where the effective gravity
478: points in the direction of the
479: less dense medium.
480: In the case of an expanding supernova
481: shockfront, the effective
482: gravity in the frame of the shockfront
483: is the shockfront acceleration.
484: Our results for the trajectory are
485: all concave down implying that
486: effective gravity points radially outward
487: and thus, away from the relatively rarer medium.
488: Hence, the shockfront is Rayleigh-Taylor stable.
489:
490: We conclude that ram pressure
491: from a collapsing cloud reduces
492: the shockfront velocity of one
493: radiative supernova explosion (occurring
494: in a localized region when the ambient
495: cloud has collapsed by 50\%) to zero at a
496: normalized radius of $1.33$ and a
497: normalized time of $0.23$.
498: These
499: fractions scale by the energy of
500: the supernova and the density of
501: the interstellar medium.
502:
503: Observations of nonthermal radio emission in galaxies
504: suggest whether or not the supernovae are confined. For
505: example, a 22 GHz map of M82 extending over a few
506: hundred parsecs is estimated to have 40 unconfined supernovae
507: (Golla et al 1996). An 18 cm nonthermal radio emission map of VIIZw19
508: extending over 310 pc is estimated to have 2500 unconfined supernovae remnants
509: (Beck. et al 2004). Two examples of apparently confined emission, SBS-0335 and Henize 2-10, are described below.
510:
511: In a model of the
512: nonthermal radio emission from SBS-0335 the emission is
513: confined to $17$ pc (Hunt et al. 2004)
514: within a $520$ pc region
515: (Thuan et al. 1997) where star
516: formation occurs in six super-star
517: clusters with ages $\leq 25$ Myrs.
518: If a supernova explosion of
519: energy $10^{51}$ ergs occurs
520: after a cloud of initial density 10$^2$ H/cm$^{3}$
521: has collapsed by 50\%,it is confined to
522: $\sim 15$ pc in $\sim 1$ Myr.
523: If this were a supernova exploding in a stationary cloud, it would
524: expand beyond $\sim 20$ pc in as little as 0.4 Myrs. This
525: would destroy the cloud and the case of a young supernova
526: remnant in a stationary cloud is considerably less probable
527: than a supernova confined in a collapsing cloud.
528:
529: Very Large Array imaging of another BCD, Henize 2-10,
530: indicates a $<8$ pc region of $1$ mJy radio
531: sources in the central $5^{''}$ starburst region.
532: Henize 2-10 has HII regions
533: of sizes between $3$ pc and $8$ pc and densities
534: between $1500$ and $5000$ H/cm$^{3}$
535: (Kobulnicky, Johnson 1999).
536: If we hypothesize that a cloud of an average initial density of
537: $3000$ H/cm$^3$ has collapsed by $50\%$ when $10$ supernovae
538: explode, we find that
539: the supernovae are confined to
540: $\sim 7$ pc in $\sim 0.2$ Myr.
541: Thus, cloud collapse successfully confines
542: supernova explosions and can account for observed
543: compact nonthermal radio emission. This simple model
544: can help understand continuous or second/third generation
545: star formation since it suggests why the cloud is not
546: devastated by first generation supernovae.
547:
548: We thank David Chernoff, Ruizhen Tan,
549: and Setu Mohta for their valuable comments.
550: Also, we thank the referee for helpful
551: suggestions. Support
552: for this work was provided by NASA through contract
553: 1257184 issued by JPL/Caltech. The work of RVEL was
554: supported in part by NASA grants NAG 5-13060,
555: NAG 5-13220, and NSF grant AST-0307817.
556:
557:
558: \begin{references}
559:
560: \reference{} Beck, S.C., Garrington, S.T., Turner, J.L.,
561: \& Van Dyk, S.D., 2004, AJ, 128, 1552
562:
563: \reference{} Bisnovatyi-Kogan, G.S.
564: \& Silich, S.A., 1995, RvMP, 67, 661
565:
566: \reference{} Cioffi, D.F.,
567: McKee, C.F. \& Bertschinger, E.
568: 1988, ApJ, 334, 252
569:
570: \reference{} Dopita, M.A., 1981, ApJ, 246, 65
571:
572: \reference{} Dopita, M.A., Ford, V.L., McGregor, P.J., Mathewson, D.S.,
573: \& Wilson, I.R., 1981, ApJ, 250, 103
574:
575: \reference{} Golla, G., Allen, M.L. \& Kronberg, P.P., 1996, ApJ, 473, 244
576:
577: \reference{} Hunt, L.K., Dyer, K.K.,
578: Thuan, T.X. \& Ulvestad,
579: J.S. 2004, ApJ, 606, 853
580:
581: \reference{} Kobulnicky, H.A. \&
582: Johnson, K.E. 1999, AAS, 31,
583: 1382
584:
585: \reference{} Salpeter, E.E., 1955, ApJ, 121, 161
586:
587: \reference{} Sedov, L.I., 1946,
588: Dokl. Akad. Nauk SSSR 42, 17
589:
590: \reference{} Taylor, G.I., 1950,
591: Proc. R. Soc. London, Ser. A 201, N1065, 159
592:
593: \reference{} Thuan, T.X., Izotov,
594: Y.I. \& Lipovetsky, V.A. 1997, ApJ, 477, 661
595:
596: \end{references}
597:
598: \clearpage
599: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
600: \begin{figure*}[h]
601: %\begin{inlinefigure}
602: %%\centerline{\epsfig{file=Fig1.eps,height=2.2in,
603: %width=3.3in, angle=0}}
604: \epsscale{1}
605: \plotone{f1.eps}
606: \figcaption{The collapse of cloud
607: shells with different initial radii
608: are shown as solid lines.
609: The radius is normalized by the
610: initial radius of the outermost cloud
611: shell and the time is normalized by
612: the free-fall time.
613: The dotted line
614: indicates the approximation of
615: the trajectory used here. The
616: approximation has a maximum deviation
617: of 5.3\% and a mean deviation of
618: 2.6\%.}
619: \end{figure*}
620: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
621:
622: \clearpage
623: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
624: \begin{figure*}[h]
625: %\begin{inlinefigure}
626: %\centerline{\epsfig{file=plot2.ps,height=3.3in,
627: %width=2.2in, angle=90}}
628: \plotone{f2.eps}
629:
630: \epsscale{1}
631: \figcaption{Shock wave expansion $R_{sh}(\tau)$
632: in a collapsing cloud and
633: the standard Sedov-Taylor
634: expansion in a stationary medium.
635: Time $\tau$ is normalized by the
636: free-fall time and the radius is
637: normalized by the Sedov-Taylor radius
638: at the free-fall time.
639: The shock expansion is slowed
640: down and subequently turned around by the collapsing cloud.
641: }
642: \end{figure*}
643: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
644:
645: \clearpage
646: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
647: \begin{figure*}[h]
648: %\begin{inlinefigure}
649: %\centerline{\epsfig{file=plot3.ps,height=3.3in,
650: %width=2.2in, angle=90}}
651: \plotone{f3.eps}
652: %\plotone{plot3.ps}
653: \epsscale{1}
654: \figcaption{Shockfront radii
655: including influence of the radiative losses.
656: We also show the shockfront radius for a time
657: lag of 50\% and 80\% of the free-fall
658: time between supernova explosion and cloud collapse.
659: Time zero is taken as the instant the supernova
660: exploded.}
661: \end{figure*}
662: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
663:
664: \clearpage
665: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
666: \begin{figure*}[h]
667: %\begin{inlinefigure}
668: %\centerline{\epsfig{file=plot4.ps,height=3.3in,
669: %width=2.2in, angle=90}}
670: \plotone{f4.eps}
671: %\plotone{plot4.ps}
672: \epsscale{1}
673: \figcaption{Evolution
674: of the shockfront radius for
675: single and multiple SN explosions.}
676: \end{figure*}
677: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
678:
679: \end{document}
680:
681: