1: \documentclass[twocolumn,a4paper]{nveart}
2: \usepackage{epsfig,amssymb,pifont}
3:
4: \begin{document}
5:
6: \company{Utrecht University}
7: \journal{}
8:
9: \begin{frontmatter}
10: \title{On the Use of Transverse Momentum Spectra to probe Thermalisation\\
11: and Collective Expansion in Relativistic Nucleus-Nucleus Collisions}
12:
13: \author{Nick van Eijndhoven}
14: \address{Department of Subatomic Physics, Utrecht University/NIKHEF\\
15: P.O. Box 80.000, NL-3508 TA Utrecht, The Netherlands}
16:
17: \begin{abstract}
18: Transverse momentum spectra of the reaction products from proton-proton and nucleus-nucleus
19: collisions are discussed within the phenomenological frameworks of statistical phase-space
20: and string fragmentation models.
21: It will be shown that a string fragmentation model can accommodate all of the observations
22: discussed here and consequently it is claimed that no conclusion about the thermal nature
23: or collective expansion of the created system may be derived from observed transverse momentum spectra.
24: \end{abstract}
25:
26: \begin{keyword}
27: Nucleus-nucleus collisions,
28: thermalisation,
29: transverse momentum spectra,
30: string fragmentation.
31: \PACS{13.75C, 13.85, 25.70.G, 24.10.P, 25.75.D}
32: \end{keyword}
33: \end{frontmatter}
34:
35: \section{Introduction}
36: With the highly-relativistic heavy-ion physics programmes reaching maturity,
37: one can now start to address the question of
38: what are the observables to provide unambiguous evidence for the formation of
39: a Quark-Gluon Plasma (QGP) state.\\
40: Before answering this question, physicists have first tried to see
41: whether the current data indicate the formation of a thermalised system,
42: the logical implication of QGP formation.
43: In doing so, transverse momenta of the particles produced in the collisions
44: have been investigated.
45:
46: In very energetic heavy-ion collisions a large number of particles is produced
47: in a relatively small volume and consequently many secondary collisions between the
48: reaction products take place (rescattering).
49: This makes experiments with highly relativistic heavy-ion beams well suited to search
50: for the formation of a thermally equilibrated system.
51:
52: Several attemps \cite{becattini,pbm} have been made to investigate whether the spectra
53: and (relative) yields of the produced particles indicate the formation of a thermally
54: and/or chemically equilibrated system.\\
55: In addition, interpretation of the observed spectra within the framework of hydrodynamical
56: evolution of the created system followed by chemical and thermal freeze-out of the hadron source
57: has led to the idea of collective expansion effects reflected by the produced particles
58: \cite{heinz,na49qm97}.
59:
60: In this paper it will be shown that particle production through the mechanism of string
61: fragmentation also leads to particle spectra compatible with the observations discussed here,
62: without the necessity of collective effects.
63:
64: \section{The thermal model}
65: A theoretical description of the momentum spectra of particles emerging from a thermalised system
66: is obtained by assuming that all the constituents obey a Boltzmann distribution.\\
67: In the case of formation of such a thermalised system by energetic collisions of heavy ions,
68: it is convenient to introduce the perpendicular and parallel momentum components with respect
69: to the direction of the incoming projectile.
70: These observables are indicated by $p_{\perp}$ and $p_{\parallel}$, respectively, whereas we
71: also introduce the transverse mass $m_{\perp} \equiv \sqrt{p_{\perp}^{2}+m^{2}}$.
72: The distribution of the emerging particles as a function of these variables is given by
73: %
74: \begin{equation}
75: {\rm d}N(\vec{p},m,T) = A\,e^{-\sqrt{m_{\perp}^{2}+p_{\parallel}^{2}}/T}m_{\perp}
76: \,{\rm d}m_{\perp}{\rm d}p_{\parallel}{\rm d}\varphi ~,
77: \label{eq:boltzmann}
78: \end{equation}
79: %
80: where $T$ represents the temperature of the system and $A$ contains the normalisation to the total
81: number $N$ of particles as well as the (mass) terms which account for the proper dimensions.
82:
83: Using the azimuthal symmetry of the situation described above,
84: integration over $p_{\parallel}$ and $\varphi$
85: results in the following expression for the differential particle yields
86: %
87: \begin{equation}
88: \frac{{\rm d}N(m_{\perp},T)}{m_{\perp}{\rm d}m_{\perp}} \sim m_{\perp}\,
89: K_{1}\left( \frac{m_{\perp}}{T} \right) ~,
90: \label{eq:thermal1}
91: \end{equation}
92: %
93: where $K_{1}$ represents the modified Bessel function.\\
94: Making use of the relations $\displaystyle K_{1}(x \gg 1) \sim \sqrt{\frac{\pi}{2x}}\,e^{-x}$
95: and $K_{1}(x \ll 1) \sim x^{-1}$ we see from Eq.~(\ref{eq:thermal1}) that
96: at high momenta an essentially exponential behaviour of the particle distribution is
97: obtained, whereas at low momenta the differential yield saturates at a level proportional
98: to the temperature of the system.\\
99: The theoretical particle spectrum for a thermalised system at a temperature of 200~MeV
100: is shown in Fig.~\ref{fig:thermal1}.
101: %
102: \begin{figure}[htb]
103: \begin{center}
104: \resizebox{8.4cm}{!}{\includegraphics{thmt.eps}}
105: \end{center}
106: \caption{Calculated particle yield for a thermal source of 200~MeV temperature
107: according to Eq.~(\ref{eq:thermal1}).}
108: \label{fig:thermal1}
109: \end{figure}
110:
111: According to the theoretical description of the particle spectra produced by a
112: thermalised system, the following observations have been made~:
113: %
114: \begin{itemize}
115: \item Various particle species will show the same shape in the transverse mass
116: spectra, the so-called $m_{\perp}$ scaling.
117: \item At high $p_{\perp}$ initial hard scattering of the constituents starts to play a role \cite{Cronin}
118: and as such an enhancement at high $p_{\perp}$ is expected, the so-called Cronin effect.
119: \item At low $p_{\perp}$ one starts to become sensitive to the products of resonance decays,
120: which induces a rapidity--dependent enhancement \cite{resonances}.
121: \end{itemize}
122: %
123: From this it is seen that, in comparison with experimental results, the description of the
124: particle spectra as outlined above is only valid in a limited $m_{\perp}$ interval.
125:
126: Furthermore, in general Eq.~(\ref{eq:thermal1}) can not be used in comparison with experimental
127: data because of the fact that, due to the limited phase-space coverage of the experimental
128: apparatus, only a limited range in $p_{\parallel}$ can be addressed.\\
129: To account for this effect, it is more convenient to express the particle yields in terms
130: of the particle rapidity
131: %
132: \begin{equation}
133: y=0.5 \ln \left( \frac{E+p_{\parallel}}{E-p_{\parallel}} \right) ~,
134: \label{eq:rapidity}
135: \end{equation}
136: %
137: where $E$ represents the particle energy.\\
138: Using the identity $p_{\parallel}=m_{\perp} \sinh(y)$ the Boltzmann distribution
139: of Eq.~(\ref{eq:boltzmann}) can be written as~:
140: %
141: \begin{equation}
142: {\rm d}N(\vec{p},m,T) \sim
143: e^{-m_{\perp}\cosh(y)/T}m_{\perp}^{2}\cosh(y)\,{\rm d}m_{\perp}{\rm d}y{\rm d}\varphi .
144: \label{eq:boltzmann2}
145: \end{equation}
146: %
147: Here it should be noted that $y$ represents the rapidity in the restframe of the thermalised
148: system, since all the descriptions so far have been performed in this frame.
149:
150: Unfortunately, the particle distribution of Eq.~(\ref{eq:boltzmann2}) can not be evaluated
151: analytically to provide differential particle yields as a function of $m_{\perp}$ and $T$
152: within a certain rapidity range.
153: For each experimental situation numerical integration over the corresponding rapidity
154: coverage has to be performed in order to compare the model with the experimental data.
155:
156: \section{What do the data tell us ?}
157: In this section the theoretical model described before will be tested against
158: experimental data.
159: For this purpose a selection of data obtained from $^{208}$Pb+$^{208}$Pb collisions at the CERN-SPS
160: with a beam energy of 158$A$~GeV and p+p collisions at the CERN-ISR at $\sqrt{s}$ of 23~GeV
161: has been used.\\
162: Here it should be realised that the $\sqrt{s}$ per nucleon-nucleon collision in case of the
163: CERN-SPS heavy-ion collisions is comparable to that of the CERN-ISR proton-proton collisions.
164:
165: Concerning the CERN-SPS heavy-ion data we will investigate the charged hadron yields as
166: observed in the NA49 experiment \cite{na49} and the $\pi^{0}$ yields as observed
167: in the WA98 experiment \cite{wa98}. For both experiments the mid-rapidity value in the
168: laboratory system was 2.9.\\
169: The various spectra are shown in Figs.~\ref{fig:na49} and \ref{fig:wa98}, respectively.
170: The fact that the spectrum of Fig.~\ref{fig:wa98} can also be compared directly with
171: Eq.~(\ref{eq:boltzmann2}) is easily demonstrated by using the identities $E=m_{\perp}\cosh(y)$
172: and ${\rm d}p_{\parallel}=E{\rm d}y$.
173:
174: \begin{figure}[htb]
175: \begin{center}
176: \resizebox{8.5cm}{!}{\includegraphics{na49.eps}}
177: \end{center}
178: \caption{Charged hadron yields \cite{na49} as observed within the NA49 experiment.
179: The solid line indicates a fit with the thermal model from Eq.~(\ref{eq:boltzmann2}).}
180: \label{fig:na49}
181: \end{figure}
182:
183: \begin{figure}[htb]
184: \begin{center}
185: \resizebox{8.5cm}{!}{\includegraphics{wa98.eps}}
186: \end{center}
187: \caption{WA98 observed $\pi^{0}$ transverse mass spectrum \cite{wa98}.
188: The solid line indicates a fit with the thermal model from Eq.~(\ref{eq:boltzmann2}).}
189: \label{fig:wa98}
190: \end{figure}
191:
192: Comparison of the observed spectra with our thermal model calculations reflected by Eq.~(\ref{eq:boltzmann2})
193: leads to the preliminary conclusion that indeed the spectra exhibit a thermal character compatible
194: with a temperature of $205$~MeV.
195: Furthermore, it has been observed \cite{na49qm97,na49} that the spectra of various particle species
196: exhibit larger inverse slopes with increasing particle mass. This phenomenon has been interpreted
197: in terms of a collective expansion of the hadronic source \cite{heinz,na49qm97,na49}. \\
198: But is it justified, based on these observations, to conclude that the observed spectra are the
199: result of a thermalised system and that the difference in (inverse) slopes for the various
200: particle species is a consequence of collective behaviour~?
201:
202: To obtain an answer to the above questions, the charged pion and kaon yields at mid-rapidity as observed
203: in proton-proton collisions at the CERN-ISR \cite{isr} have been investigated as is
204: shown in Fig.~\ref{fig:isr}.
205:
206: \begin{figure}[htb]
207: \begin{center}
208: \resizebox{8.5cm}{!}{\includegraphics{isrkpip.eps}}
209: \end{center}
210: \caption{Charged pion ({\tiny \ding{108}}) and kaon ({\tiny \ding{110}}) yields observed in
211: p-p collisions at $\sqrt{s}=23$~GeV at the CERN-ISR \cite{isr}.
212: The solid lines indicate fits with the thermal model from Eq.~(\ref{eq:boltzmann2}).}
213: \label{fig:isr}
214: \end{figure}
215:
216: From the CERN-ISR data it is seen that the pion and kaon spectra (similar results were
217: observed for $\pi^{-}$ and $K^{-}$ spectra) produced in proton-proton reactions at
218: similar collision energies as the CERN-SPS heavy-ion data are also compatible with our
219: calculated spectrum reflected by Eq.~(\ref{eq:boltzmann2}).
220: In addition, also here it is seen that particles with larger mass exhibit a larger inverse
221: slope in the $m_{\perp}$ spectrum.
222: The temperatures obtained here are 190~MeV and 223~MeV for the pions and kaons respectively,
223: which even quantitatively agree with the observations in the case of heavy-ion collisions
224: \cite{na49qm97}.
225: It should be noted that in \cite{na49qm97} a slightly different fit was used, but in the
226: $m_{\perp}$ range considered here this leads to only marginal differences in the extracted
227: temperatures.
228:
229: Furthermore, even for high energy $e^{+}e^{-}$ annihilation reactions at the PETRA and LEP
230: accelerators \cite{lep} the particle spectra are found to match calculations based
231: on a thermal model.
232: Since it is very unlikely that in proton-proton and $e^{+}e^{-}$ collisions a thermalised
233: system will be created and collective expansion effects are present, the interpretation of the
234: observed $m_{\perp}$ spectra in terms of a thermalised and expanding hadronic source becomes
235: doubtful.\\
236: Further investigations to obtain a more detailed understanding of the underlying
237: reaction dynamics are therefore needed. Below, an attempt is made to explain the observed
238: distributions based on a common underlying production mechanism.
239:
240: \section{How to understand the observations~?}
241: Due to the fact that the interactions described here are not calculable within QCD,
242: the interpretation of experimental results has to rely on phenomenological approaches.
243:
244: \subsection{Statistical approach}
245: First attempts to interpret the experimental data of high energy nuclear collisions were performed
246: by trying to describe statistically the stage in which excitation of the incoming matter
247: takes place. However, in their original formulation \cite{stat1} these descriptions of
248: this so-called 'early stage' of the interaction failed to reproduce the experimental results.
249:
250: With the advent of more and more high quality data it turned out that after some
251: generalisations had been applied to the original ideas, a statistical approach of the
252: early stage of the interaction \cite{marek} gives rather good agreement with the experimental results.\\
253: Furthermore, recently it has been demonstrated \cite{becattini,becattini2,jean} that also the hadronisation
254: process, which follows after the expansion and cooling of the early stage, can be described successfully
255: by a statistical model.
256:
257: Provided an extrapolation to a full phase-space coverage is performed, these statistical models
258: describe successfully the various hadron abundancies (i.e. the chemical composition)
259: of the interactions as observed in the various experiments.
260: However, they provide no description of the dynamical properties of the
261: created system as for instance reflected in the $m_{\perp}$ spectra of the reaction products.\\
262: These models also fail to describe the observed event topologies like the leading particle effect
263: and jet structures in $e^{+}e^{-}$ annihilations
264: and in addition need the explicit assumption that the temperature of the system
265: remains constant during the complete hadronisation proces.
266: Furthermore, an ad-hoc strangeness suppression factor $(\gamma_{s})$ has to be introduced
267: to accommodate the observed yields of particles containing strangeness.
268:
269: All this makes one wonder whether such statistical approaches are the right way to go.
270: Perhaps a description of the interaction process on a much more elementary level could
271: provide a proper description of all the observed characteristics.
272:
273: \subsection{String fragmentation model}
274: Based on an intuitive interpretation of the Dual Resonance Model a string
275: model of hadrons has been developed \cite{string1}.
276: This model was seen to provide a simple explanation for confinement and also
277: leads to the fact that the particle spin is proportional to the square of the
278: particle mass, the so-called Regge behaviour.
279:
280: Based on the success of the string model, such a string formalism was also
281: used to provide a description of high energy collisions \cite{artru1,artru2} and
282: it was seen that such a phenomenological approach provided a quantitative
283: description of experimentally observed effects such as~:
284: %
285: \begin{itemize}
286: \item Leading particle effect and jet structures.
287: \item The existence of a rapidity plateau around mid-rapidity.
288: \item Limiting fragmentation, resulting in a logarithmic growth of the particle multiplicity
289: as a function of the collision energy.
290: \item Relatively small values of the transverse momentum of the produced particles.
291: \end{itemize}
292: %
293: Independently of the classical string phenomenology, a QCD inspired string fragmentation
294: model was developed \cite{lund} which also provided a quantitative description of
295: high energy interactions.
296: The latter description has the advantage that it is well suited to be implemented
297: in computer simulation programs, the so-called Lund Monte Carlo program.\\[5mm]
298: Both scenarios turn out to give similar results and in this paper the Lund
299: formalism will be used.
300:
301: \subsubsection{Transverse momentum spectra}
302: In section~5 of \cite{lund} it is demonstrated that fragmentation of a stretched-out
303: colour field provides a common mechanism to successfully describe particle production
304: for all of the following processes~:
305: %
306: \begin{itemize}
307: \item Quark fragmentation in $e^{+}e^{-}$ annihilations.
308: \item Meson fragmentation in hadronic collisions.
309: \item Proton fragmentation in hadronic collisions.
310: \item Proton fragmentation in deep inelastic scattering\\
311: (diquark fragmentation).
312: \end{itemize}
313: %
314: In view of the similarities concerning the various $m_{\perp}$ spectra mentioned before,
315: it is therefore tempting to try to explain the observed $m_{\perp}$ distributions within
316: the framework of a string fragmentation model.
317:
318: As outlined in \cite{lund}, hadron-hadron interactions can be described by representing
319: the initial hadrons as massless quarks connected by a string in "yo-yo mode" for a meson
320: and by a Y-shaped string for a baryon.
321: Due to the relative motion of the quarks in the initial hadron the colour field will not
322: be completely straight, but effectively in the overall cms of the collision a linear
323: description is obtained by projection on the direction $\vec{Q}+\vec{k}$, where $\vec{Q}$
324: represents the momentum transfer and $\vec{k}$ the primordial momentum of the struck (anti)quark.\\
325: Comparison of the string fragmentation model with various data indicates that the value of
326: $k_{\perp}$ (with respect to the direction $\vec{Q}+\vec{k}$) is of the order of about 400~MeV/c.
327: The remaining momentum $-\vec{k}$ of the system when the reaction takes place is distributed
328: among the particles in the "target" fragmentation region in which a particle with a large
329: energy-momentum fraction $z$ also takes a large fraction of this remaining momentum.
330:
331: Limiting ourselves to soft processes and light-particle production at mid-rapidity
332: (i.e. low $z$ values), it follows that the effects of the primordial $k_{\perp}$
333: are marginal in our discussion concerning the observed $m_{\perp}$ spectra mentioned before.
334: In particular for initial baryons one quark will always be close to the junction which leads
335: to an essentially linear picture describing the diquark fragmentation along the incoming
336: beam direction.
337:
338: In the following the string fragmentation model will be used to derive the $m_{\perp}$
339: spectra for mesons produced at mid-rapidity in the case of hadron fragmentation.\\
340: Consider the case where, after string fragmentation, a quark of flavour $\alpha$ and an
341: anti-quark of flavour $\beta$ combine into a meson with transverse mass $m_{\perp}$ and
342: energy-momentum fraction $z$.
343: According to \cite{lund} the probability for this process to occur in the case of symmetric
344: string fragmentation is given by the so-called scaling function $f_{\alpha\beta}(z,m_{\perp})$
345: which is found to be
346: %
347: \begin{equation}
348: f_{\alpha\beta}(z,m_{\perp})=N_{\alpha\beta} (1-z)^{a} \, \frac{1}{z} \, e^{-bm_{\perp}^{2}/z} ~,
349: \label{eq:fab}
350: \end{equation}
351: %
352: where $a$ and $b$ are free parameters and $N_{\alpha\beta}$ represents a normalisation constant.\\
353: For low $z$ values the probability reflected in Eq.~(\ref{eq:fab}) can be written as
354: %
355: \begin{equation}
356: f_{\alpha\beta}(z,m_{\perp}) \sim \frac{1}{z} \, e^{-bm_{\perp}^{2}/z} ~,
357: \label{eq:fab2}
358: \end{equation}
359: %
360: whereas for $z \rightarrow 1$ the expression of Eq.~(\ref{eq:fab}) basically becomes $(1-z)^{a}$.
361:
362: Following the phenomenology outlined in detail in section~2.6 of \cite{lund} it is seen
363: that the parameters $a$ and $b$ are strongly correlated, since increasing $a$ or decreasing
364: $b$ will increase the mean multiplicity and affect the shape of the transverse momentum spectra.
365: In order to match the experimentally observed spectra and multiplicities only values along a curve
366: in the $(a,b)$ plane are allowed.\\
367: Good agreement with various data is obtained by $a=1$ and $b=0.4$~GeV$^{-2}$.
368: Using these values of $a$ and $b$, it is seen that particle production at high $z$ values
369: is suppressed and that the distribution at low $z$ values peaks at $z=0.4 m_{\perp}^{2}$
370: (for low $m_{\perp}$ values) and vanishes for smaller $z$ values.
371:
372: As discussed before, light-particle production at mid-rapidity due to soft processes
373: involves low $z$ values.
374: If we thus limit ourselves to this part of phase-space and make use of
375: $E=m_{\perp}\cosh(y)$, $p_{\parallel}=m_{\perp}\sinh(y)$ and
376: $z^{\pm}=(E \pm p_{\parallel})_{hadron}/(E \pm p_{\parallel})_{total}$,
377: the latter of which are the invariant lightcone variables describing the "forward" and
378: "backward" fragmentation in case of finite jet energies \cite{lund},
379: it is readily seen that in our case $z \sim m_{\perp}$.
380:
381: This eventually leads to the following distribution function concerning light-particle
382: production in the central region~:
383: %
384: \begin{equation}
385: f(m_{\perp}) \sim \frac{1}{m_{\perp}} \, e^{\lambda m_{\perp}} ~,
386: \label{eq:fstring}
387: \end{equation}
388: %
389: where $\lambda$ comprises both the constant $b$ and the term $(E \pm p_{\parallel})_{total}$
390: corresponding to the forward or backward fragmentation.\\
391: From this probability the distribution of the particles emerging around mid-rapidity is found to be
392: %
393: \begin{equation}
394: {\rm d}N(\vec{p},m) \sim
395: e^{\lambda m_{\perp}}m_{\perp}\cosh(y)\,{\rm d}m_{\perp}{\rm d}y{\rm d}\varphi .
396: \label{eq:string}
397: \end{equation}
398: %
399: Here it is seen that also the phenomenological string fragmentation model results in
400: a transverse mass distribution of the particles produced in high energy collisions
401: which exhibits a shape compatible with a thermal source.\\
402: Obviously one could now raise the question of how Eq.~(\ref{eq:string}) would fit the
403: experimental spectra.
404:
405: Taking again the charged pion and kaon yields from the CERN-ISR proton-proton data as shown
406: in Fig.~\ref{fig:isr} and performing a fit of the data with Eq.~(\ref{eq:string}) shows that
407: the model describes the data rather well as can be seen from Fig.~\ref{fig:stringfit}.
408:
409: \begin{figure}[htb]
410: \begin{center}
411: \resizebox{8.5cm}{!}{\includegraphics{stringfit_isr.eps}}
412: \end{center}
413: \caption{Charged pion ({\tiny \ding{108}}) and kaon ({\tiny \ding{110}}) yields observed in
414: p-p collisions at $\sqrt{s}=23$~GeV at the CERN-ISR \cite{isr} fitted against
415: the string fragmentation model from Eq.~(\ref{eq:string}).}
416: \label{fig:stringfit}
417: \end{figure}
418: %
419: Similar results are obtained for the $\pi^{-}$ and $K^{-}$ data.
420: The inverse slopes of the fit, which in a thermal model would represent the
421: temperature of the system, are found to be 218~MeV and 257~MeV for the pions and kaons, respectively.
422: As outlined in section 2.6 of \cite{lund}, also from the string fragmentation model it is seen that
423: the inverse slope becomes larger for heavier particles since they need a slightly higher
424: energy-momentum fraction $z$ at their creation.
425: Considering momentum values in the range of 0.5-1 GeV/c it is seen that kaons exhibit an increase
426: in $z$ of about 20\% compared to pions, resulting in a comparable increase in inverse slope of
427: the $m_{\perp}$ spectra.
428: This is perfectly in line with the observations in the case of the ISR data discussed above
429: and also in the case of heavy-ion collisions at the CERN-SPS \cite{na49qm97}.\\
430: Here it should be noted that the production of particles containing strangeness needs a
431: separate study concerning heavy-quark production in the string fragmentation model.
432: Since this is outside the scope of this paper we will not elaborate on this subject here,
433: but it is seen \cite{lund} that, apart from a relative suppression factor of about $\frac{1}{3}$
434: for strange-quark production, the fragmentation into kaons proceeds similarly to that into pions.\\
435: It is thus seen that the string model also automatically yields a suppression of heavy-quark production
436: which turns out to be in agreement with experimental observations and even matches the ad-hoc
437: strangeness suppression factor $\gamma_s=0.3$ which had to be introduced in the statistical
438: model \cite{becattini,becattini2,jean}.
439:
440: Comparison of the nucleus-nucleus data from Figs.~\ref{fig:na49} and \ref{fig:wa98} with the
441: string fragmentation model of Eq.~(\ref{eq:string}) is presented in the Figs.~\ref{fig:stringfit_na49}
442: and \ref{fig:stringfit_wa98} below.
443:
444: \begin{figure}[htb]
445: \begin{center}
446: \resizebox{8.5cm}{!}{\includegraphics{stringfit_na49.eps}}
447: \end{center}
448: \caption{Charged hadron yields \cite{na49} as observed within the NA49 experiment.
449: The solid line indicates a fit with the string fragmentation model from Eq.~(\ref{eq:string}).}
450: \label{fig:stringfit_na49}
451: \end{figure}
452:
453: \begin{figure}[htb]
454: \begin{center}
455: \resizebox{8.5cm}{!}{\includegraphics{stringfit_wa98.eps}}
456: \end{center}
457: \caption{WA98 observed $\pi^{0}$ transverse mass spectrum \cite{wa98}.
458: The solid line indicates a fit with the string fragmentation model from Eq.~(\ref{eq:string}).}
459: \label{fig:stringfit_wa98}
460: \end{figure}
461:
462: Also in the case of the $m_{\perp}$ spectra of the hadrons produced in nucleus-nucleus
463: collisions the string fragmentation model is seen to describe the data rather well.
464: The inverse slopes from the NA49 and WA98 data are found to be 237~MeV and 227~MeV respectively.\\
465: In order to compare the various observed (inverse) slopes, the values obtained from the NA49
466: and WA98 data have to be corrected for the rapidity coverage.
467: Using the values of the specified rapidity intervals expressed in the overall cms of the collision
468: and following the arguments given above, both inverse slopes amount to about 200~MeV after
469: the rapidity correction.\\
470: From this it is seen that indeed the string fragmentation model is able to describe all the
471: data discussed in this paper in a consistent way.
472:
473: \section{Conclusions}
474: Various models describing particle production were confronted with data from
475: proton-proton and nucleus-nucleus data.\\
476: It was seen that both a thermal model and a phenomenological string fragmentation
477: model describe the observed transverse momentum spectra, where the string fragmentation
478: model automatically comprises the increase in inverse slope with particle mass as has been
479: observed in both the p-p and heavy-ion data discussed here.
480: To accommodate this in the thermal model, different freeze-out temperatures have to be
481: assumed for the various particle species or a collective expansion of the system has to be imposed,
482: in which the expansion velocity is the same for all particle species.
483:
484: Concerning the production of various quark flavours, the string fragmentation model was seen
485: to yield automatically suppression of heavy quarks as compared to the light $u$ and $d$ flavours.
486: The strangeness suppression factor of about $\frac{1}{3}$ turned out to match the ad-hoc introduced
487: $\gamma_{s}$ from statistical models.\\
488: Finally it should be noted that the string model is the only one which accounts for the
489: observed event topologies like for instance the leading particle effect and jet structures.
490:
491: From this it is concluded that a phenomenological string fragmentation model describes
492: all aspects of the data discussed here and consequently it is not justified to draw any
493: conclusions concerning a thermal nature or a collective expansion of the created system from
494: the observed transverse momentum spectra.
495:
496: \begin{ack}
497: The author would like to thank Michiel Botje, Adriaan Buijs and Marco van Leeuwen
498: for the inspiring discussions and valuable suggestions.\\
499: He would also like to acknowledge the ROOT team for providing such a powerful
500: analysis framework.
501: \end{ack}
502:
503: \begin{thebibliography}{99}
504: \bibitem{becattini}
505: F. Becattini, {\em et al.}, Nucl. Phys. {\bf B71} (1999) 324.
506: \bibitem{pbm}
507: P. Braun-Munzinger, {\em et al.}, Phys. Lett. {\bf B465} (1999) 15.
508: \bibitem{heinz}
509: J. Sollfrank, {\em et al.}, Z. Phys. {\bf C52} (1991) 593.\\
510: E. Schnedermann, {\em et al.}, Phys. Rev. {\bf C48} (1993) 2462.\\
511: U. Heinz, Nucl. Phys. {\bf A610} (1996) 264.
512: \bibitem{na49qm97}
513: {\em NA49 collaboration}, Nucl. Phys. {\bf A638} (1998) 91c.
514: \bibitem{Cronin}
515: D. Antreasyan, {\em et al.}, Phys. Rev. {\bf D19} (1979) 764.
516: \bibitem{resonances}
517: L. Ahle, {\em et al.}, Nucl. Phys. {\bf A610} (1996) 139c.
518: \bibitem{na49}
519: {\em NA49 collaboration}, Phys. Rev. Lett. {\bf 82} (1999) 2471.
520: \bibitem{wa98}
521: {\em WA98 collaboration}, Phys. Rev. Lett. {\bf 81} (1998) 4087.
522: \bibitem{isr}
523: B. Alper, {\em et al.}, Nucl. Phys. {\bf B87} (1975) 19.\\
524: K. Guettler, {\em et al.}, Phys. Lett. {\bf B64} (1976) 111.
525: \bibitem{lep}
526: H.G. Sander, Nucl. Phys. {\bf B71} (1999) 123.
527: \bibitem{stat1}
528: E. Fermi, Prog. Theor. Phys. {\bf 5} (1950) 570.
529: \bibitem{marek}
530: M. Gazdzicki, Z. Phys. {\bf C65} (1995) 215.
531: \bibitem{becattini2}
532: F. Becattini, Z. Phys. {\bf C69} (1996) 485.
533: \bibitem{jean}
534: J. Cleymans and K. Redlich, Phys. Rev. Lett. {\bf 81} (1998) 5284.
535: \bibitem{string1}
536: L. Susskind, Phys. Rev. {\bf D1} (1970) 1182.
537: \bibitem{artru1}
538: X. Artru and G. Mennessier, Nucl. Phys. {\bf B70} (1974) 93.
539: \bibitem{artru2}
540: X. Artru, Phys. Rep. {\bf 97} (1983) 147.
541: \bibitem{lund}
542: B. Andersson, {\em et al.}, Phys. Rep. {\bf 97} (1983) 33.
543: \end{thebibliography}
544:
545: \end{document}
546: