Finite time pseudo-rip singularity in cosmology Mariusz P. Da̧browski, Teodor Borislavov Vasilev Abstract By studying first a new decelerating sudden future singularity (SFS) universe we report finding a novel type of cosmological singularity which we dub a finite time pseudo-rip (FTPR) because unlike for a pseudo-rip, it happens in the finite future of the universe. In contrast to the new SFS model, where the expansion is decelerating before reaching the pressure singularity, the FTPR scenario is preceded by a super-accelerated phantom phase. Our claim is based on the thorough study of the energy conditions showing the violations of all of them for a FTPR, and only the dominant energy one for an SFS. Application of the so-called Raychaudhuri averaging shows that, alike within the requirement of geodesic completeness, these singularities are weak in the sense of this definition. We study the properties of the models including the behaviour of the cosmological horizons presented in the appropriate Penrose diagrams. I Introduction Since the discovery of the accelerated expansion of the Universe and the postulation of dark energy Riess et al. 1998 ; Perlmutter et al. 1999 , together with recent reports of observational tensions Di Valentino et al. 2021 , a variety of models have been proposed as alternatives to the standard \Lambda CDM cosmology. These models aim to provide a viable late-time behaviour while alleviating some of the conceptual problems of the cosmological constant Weinberg 1989 ; Adler et al. 1995 . However, deviations from the standard model are usually accompanied with theoretical and observational challenges of their own Bull et al. 2016 ; Perivolaropoulos and Skara 2022 . In particular, the exploration of strongly negative pressure components, such as phantom energy Caldwell 2002 ; Caldwell et al. 2003 ; Da̧browski et al. 2003 , has revealed the possibility of future singularities like the big-rip (BR), leading to the destruction of all bound structures, and has motivated numerous extensions that fully or partially violate the Hawking–Penrose energy conditions (EC) in General Relativity Hawking and Ellis 1973 . These singularities showed up consecutively following Barrow’s paper on the sudden future singularity (SFS) models Barrow 2004a ; Barrow 2004b and later have been developed by other researches. The first summary of the possibilities called type I to IV have been given in Ref. Nojiri et al. 2005 and later expanded, among others, in Ref. Da̧browski 2014 . In general, models which fall outside the classical big-bang (BB) singularity are also known under the name of “exotic” singularities due to their diverse behaviour in comparison to their original canon. In this paper, we investigate two new exotic singularity models by postulating some specific behaviour of the Hubble parameter related to the cosmological fluid energy density: one violating the dominant energy condition only and another violating all the energy conditions. Both models lead to a future curvature singularity in a finite value of cosmological time. These singularities may be classified as an SFS and a new type of event, hereafter referred to as a finite-time pseudo-rip (FTPR), which–up to certain subtleties discussed in the paper–differs from other intrinsic phantom scenarios such as the pseudo-rip (PR) Frampton et al. 2012 ; Trivedi and Timoshkin 2024 , the little sibling of the big-rip (LSBR) Bouhmadi-López et al. 2015 , the little-rip (LR) Frampton et al. 2011 , the BR Caldwell 2002 ; Caldwell et al. 2003 ; Da̧browski et al. 2003 or the big-freeze (BF) Bouhmadi-López et al. 2008 ; Bouhmadi-López et al. 2008 . Our paper is organized as follows. In Section II we present our universe models which admit the exotic singularities. In Section III we explore these models in terms of the energy conditions. In Section IV we present the Penrose diagrams for the models which allow to give insight into their asymptotic structure. In Section V we test the models using the Raychaudhuri weak singularity criterion. In the last Section VI we present our conclusions. II New exotic singularity models II.1 A decelerating sudden future singularity model We propose a new model of SFS Barrow 2004a (or type II Nojiri et al. 2005 ; Gorini et al. 2004 ; Da̧browski et al. 2007 ; Keresztes et al. 2009 ; Keresztes et al. 2010 ) universe which is described by the Hubble rate \displaystyle H[a(t)]\equiv\frac{\dot{a}}{a}=H_{\star}\frac{a_{s}^{n}}{a^{2}(t)}\left(1-\frac{a(t)}{a_{s}}\right)^{n}, (1) where a(t) is the scale factor, and H_{\star} and a_{s} are positive constants. We also consider that n is a positive parameter that obeys the restriction 00 , says constant for n=0 or diverge for n<0 . The acceleration of the scale factor ( 3 ) vanishes for n>1/2 , converges to a constant for n=1/2 or diverges to infinity for n<1/2 . Thus, for n\in(0,1/2) the effective energy density vanishes whereas the effective pressure diverges at a_{s} . As advanced in expression ( 2 ), this behaviour correspond to a SFS or Type II singularity; see, for instance, Nojiri et al. 2005 . Interestingly, the pressure starts from infinity at BB, then diminishes reaching minimum in order to blow up to (positive) infinity again at SFS (cf. Fig. 1 ). This signals a decelerating behaviour near the SFS. Nevertheless, the equation ( 3 ) confirms that the model at hands represents, in fact, an ever decelerating universe since \ddot{a} is always negative for a\in(0,a_{s}) . In our approach, SFS plays the role of an extra fluid in the universe apart from the radiation. Figure 1 : The evolution of the decelerating model ( 1 ) for n=0.2 . Left panel: the late-time evolution of the pressure p given by ( 6 ). Right panel: the evolution of the equation of state parameter w according to ( 6 ). Figure 2 : The evolution of the decelerating scale factor ( 7 ) in time, normalized by the appropriate values at SFS (here n=0.2 ). The dependence of the scale factor a on the cosmological time, t , can easily be obtained by integrating ( 1 ), and reads \displaystyle 1-\left(1-\frac{a}{a_{s}}\right)^{1-n}\left(1-(n-1)\frac{a}{a_{s}}\right)=\frac{t}{t_{s}}, (7) where \displaystyle t_{s}\coloneqq\frac{a_{s}^{2-n}}{(n-1)(n-2)H_{\star}}, (8) is the time of SFS occurrence. An inverse of ( 7 ), i.e. the solution of the scale factor as a function of the cosmological time, a=a(t) , is not possible in terms of elementary functions when the parameter n is arbitrary. This is why, in Fig. 2 , we solve the time evolution of the scale factor numerically for n=0.2 as an example for the behaviour of our model ( 1 ). Two physically relevant regimes at which the solution of ( 7 ) is indeed possible in terms of elementary functions are when: ( i ) a\approx 0 is small (that is near the BB singularity), ( ii ) for a\approx a_{s} close to the exotic singularity. In the former case, by expanding ( 7 ) in series of small a/a_{s} , it is possible to obtain that \frac{a(t)}{a_{s}}\approx\sqrt{\frac{2}{(n-1)(n-2)}}\left(\frac{t}{t_{s}}\right)^{1/2}, (9) as it should be for the radiation-dominated universe. For the later case, i.e. when close to the SFS, the scale factor reads \displaystyle\frac{a(t)}{a_{s}}\approx 1-\left(\frac{1-\frac{t}{t_{s}}}{2-n}\right)^{\frac{1}{1-n}}. (10) It is also useful to calculate the conformal time \eta which reads \displaystyle\eta \displaystyle=\int\frac{dt}{a(t)}=\int\frac{da}{a^{2}H(a)} \displaystyle=-\eta_{0}\left(1-\frac{a}{a_{s}}\right)^{1-n}, (11) where \displaystyle\eta_{0}\coloneqq\frac{a_{s}^{1-n}}{(1-n)H_{\star}}, (12) is an integration constant. Please note that for the model ( 1 ), we have an initial BB at t=0 followed by a SFS at t=t_{s} . In terms of the conformal time this implies that \eta goes from -\eta_{0} at the BB to 0 at the SFS. Using the conformal time, we introduce the particle horizon (PH) and the event horizon (EH) as Kanai et al. 2023 \displaystyle r_{PH}(\eta)= \displaystyle\,\eta+\eta_{0}, (13) \displaystyle r_{EH}(\eta)= \displaystyle-\eta. (14) Besides, the apparent horizon (AP) in the conformal time reads Kanai et al. 2023 \displaystyle r_{AH}(\eta)\coloneqq\frac{1}{a(\eta)\,H(\eta)}, (15) see, for instance, Ref. Melia 2018 and references therein. We will use these horizons in Section IV for discussing the causal structure of the corresponding FLRW spacetime. II.2 An accelerating towards a pseudo-rip singularity model The new model ( 1 ) parametrise an ever decelerating universe which lies at total odds with current observations. Late-time acceleration which would mimic the dark energy due to a new singularity appearance can be achieved by modifying the model ( 1 ). An option is to make the terms responsible for BB domination ( \sim a^{-2} ) and SFS domination (\sim(1-a/a_{n})^{n}) additive, not multiplicative. This is \displaystyle H[a(t)]\equiv\frac{\dot{a}}{a}=\frac{H_{1}}{a^{2}}-H_{2}a_{s}^{n}\left(1-\frac{a}{a_{s}}\right)^{n}, (16) where a minus sign between the two terms is necessary to produce the acceleration. The second derivative of the scale factor a(t) , follows from ( 16 ), and reads \displaystyle\frac{\ddot{a}}{a} \displaystyle= \displaystyle-\frac{H_{1}^{2}}{a^{4}}+nH_{1}H_{2}\frac{a_{s}^{n-1}}{a}\left(1-\frac{a}{a_{s}}\right)^{n-1} (17) \displaystyle+ \displaystyle H_{2}^{2}a_{s}^{2n}\left(1-\frac{a}{a_{s}}\right)^{2n}-nH_{2}^{2}aa_{s}^{2n-1}\left(1-\frac{a}{a_{s}}\right)^{2n-1}. From ( 17 ) we notice that its first term always decelerates, the second always accelerates and eventually blows up towards minus infinity at a=a_{s} for 00 . Of course at singularity for a=a_{s} , \varrho c^{2}+p\to\infty , because of the pressure blow-up. WEC is also trivially fulfilled on the base of ( 4 ) ( \varrho\sim H^{2}\geq 0 always), and the condition ( 19 ). In order to check the expected violation of DEC at the SFS, we inspect separately the DEC-I and DEC-II. Using equations ( 1 ) and ( 3 ), the requirement to fulfil the DEC-I leads to the condition \frac{a}{a_{s}}\leq\frac{2}{2-n}, (20) which is fulfilled for all a\in[0,a_{s}] for the values of interest for the parameter n , i.e. for n\in(0,1/2) . On the contrary, the fulfilment of the DEC-II demands \frac{a}{a_{s}}\leq\frac{1}{1+n}. (21) Since 1/1+n is clearly less than one for n positive, this condition can only be satisfied at some early stages in the evolution of the model. However, when a\approx a_{s} it is always violated and, consequently, the DEC too. This is, of course, to be expected since the SFS should break the DEC condition Barrow 2004a ; Barrow 2004b . Finally, the SEC fulfilment requires \frac{a}{a_{s}}\leq\frac{1}{1-n}, (22) and it is always satisfied on the base of the condition that a\leq a_{s} and the assumption of n\in(0,1/2) . Summarising the EC for SFS models, the NEC is generally fulfilled while DEC is not since the pressure dominates at the singularity. The singularity is weak, at least in the sense of Tipler Tipler 1977 and Królak Królak 1986 , though the pressure has a finite-time divergence. So, SFS can also be called “sudden pressure singularity” since the energy density does not blow-up. III.2 Accelerating PR model For the model ( 16 ), the NEC (i.e. \dot{H}\leq 0 ) reads \displaystyle 2H_{1}H_{2}\frac{a_{s}^{n}}{a^{2}}\left(1-\frac{a}{a_{s}}\right)^{n}+nH_{1}H_{2}\frac{a_{s}^{n-1}}{a}\left(1-\frac{a}{a_{s}}\right)^{n-1} \displaystyle-2\frac{H_{1}^{2}}{a^{4}}-nH_{2}^{2}aa_{s}^{2n-1}\left(1-\frac{a}{a_{s}}\right)^{2n-1}\leq 0, (23) which shows that it is fulfilled at t=0 when a\approx 0 , i.e. at the BB, since the third term in ( III.2 ) dominates and it is negative. However, at t\approx t_{s} when a\approx a_{s} we can differentiate two cases. These are: i ) 0\leq n<1/2 , and ii ) 1/20 , r\geq 0 , and d\Omega_{3} being the line element of a 2-dimensional sphere. Introducing the null coordinates \displaystyle u= \displaystyle\eta-r, (25) \displaystyle v= \displaystyle\eta+r, (26) where u\in(-\infty,0] and v\in[-\eta_{0},\infty) with u\leq v , we may bring the line element ( 24 ) into \displaystyle ds^{2}=a(u,v)^{2}\left[-dudv+\frac{(v-u)^{2}}{4}d\Omega_{3}^{2}\right]. (27) Name t \eta r u v U V T R T_{0} 0 -\eta_{0} 0 -\eta_{0} -\eta_{0} -\arctan(\eta_{0}) -\arctan(\eta_{0}) -2\arctan(\eta_{0}) 0 i^{0} 0 -\eta_{0} \infty -\infty \infty -\frac{\pi}{2} \frac{\pi}{2} 0 \pi O t_{s} 0 0 0 0 0 0 0 0 i^{0} t_{s} 0 \infty -\infty \infty -\frac{\pi}{2} \frac{\pi}{2} 0 \pi Table 1 : Relevant points for the Penrose diagram in the different sets of coordinates used in Sec. IV . The coordinates (u,v) can be transformed into the new ones, (U,V) by the following compactification scheme \displaystyle U= \displaystyle\arctan(u), (28) \displaystyle V= \displaystyle\arctan(v), (29) which fulfil that U\in(-\pi/2,0] and V\in[-\arctan(\eta_{0}),\pi/2) , and U\leq V . In these coordinates, the line element reads \displaystyle ds^{2}=\frac{a(u,v)^{2}}{\cos^{2}U\cos^{2}V}\left[-dUdV+\frac{\sin^{2}(v-u)}{4}d\Omega_{3}^{2}\right]. (30) Finally, this line element can be mapped into the Einstein static cylinder universe by considering the following time- and space-like compactified coordinates \displaystyle T= \displaystyle U+V, (31) \displaystyle R= \displaystyle V-U. (32) Note that T\in[-T_{0},0] and R\in[0,\pi) where \displaystyle T_{0}\coloneqq-2\arctan(\eta_{0}). (33) Moreover, T_{0}>-\pi since \eta_{0} is finite. The line element now reads \displaystyle ds^{2}=\frac{a(T,R)^{2}}{\left[\cos(T)+\cos(R)\right]^{2}}\left(-dT^{2}+dR^{2}+\sin^{2}(R)d\Omega_{3}^{2}\right), (34) which is nothing, but the Einstein cylinder up-to a conformal factor. The important points for the Penrose diagrams are summarised in Table 1 . It is interesting to note that the past time-like infinity point i^{-} and null past infinity \mathcal{I}^{-} are not present in Table 1 since \eta in our models is always finite. The Penrose diagrams for the decelerating SFS ( 1 ) and accelerating FTPR model ( 16 ) can be found in Fig. 4 . In principle, both diagrams can be extended beyond the upper boundary T=0 , which corresponds to the occurrence of these singularities, since these events are geodesically extendible Fernández-Jambrina and Lazkoz 2004 . Similar diagrams for different SFS models have been previously found in Ref. Da̧browski and Marosek 2018 . Figure 4 : Comparison of the Penrose diagrams for the models discussed in this section, where EH stands for the event horizon, PH for the particle horizon, and AH is the apparent horizon. Left panel: diagram for the model ( 1 ) representing a decelerating SFS with n=0.2 . Right panel: diagram for the accelerating model ( 16 ) with n=0.2 , a_{s}=1 and H_{1}=H_{2}=1 . In addition, r_{AH}^{\star} represents the spatial radius of AH evaluated at the FTPR. For the decelerating SFS modelled by equation ( 1 ), the spatial radius of the different horizons previously mentioned obey the following constraints \displaystyle r_{EH}>r_{AH}\geq r_{PH}\ \ \textup{at BB,} (35) \displaystyle r_{AH}>r_{PH}>r_{EH}\ \ \textup{at SFS,} (36) see left panel in Fig. 4 . However, this spatial hierarchy between different horizons changes when having an accelerated regime, such as in model ( 16 ) (see right panel in Fig. 4 ). More concretely, the spatial radius for the AH grows more slowly in the accelerating FTPR universe than in the decelerating SFS one. This should not be surprising given that \displaystyle\mathchoice{\frac{\mathrm{d}\mkern 0.0mu\,r_{AH}}{\mathrm{d}\mkern 0.0mu\eta}}{\displaystyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{AH}}{\mathrm{d}\mkern 0.0mu\eta}}}{\scriptstyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{AH}}{\mathrm{d}\mkern 0.0mu\eta}}}{\scriptstyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{AH}}{\mathrm{d}\mkern 0.0mu\eta}}}=-\frac{\ddot{a}}{aH^{2}}, (37) whereas, \displaystyle\mathchoice{\frac{\mathrm{d}\mkern 0.0mu\,r_{PH}}{\mathrm{d}\mkern 0.0mu\eta}}{\displaystyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{PH}}{\mathrm{d}\mkern 0.0mu\eta}}}{\scriptstyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{PH}}{\mathrm{d}\mkern 0.0mu\eta}}}{\scriptstyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{PH}}{\mathrm{d}\mkern 0.0mu\eta}}} \displaystyle=1, (38) \displaystyle\mathchoice{\frac{\mathrm{d}\mkern 0.0mu\,r_{EH}}{\mathrm{d}\mkern 0.0mu\eta}}{\displaystyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{EH}}{\mathrm{d}\mkern 0.0mu\eta}}}{\scriptstyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{EH}}{\mathrm{d}\mkern 0.0mu\eta}}}{\scriptstyle{\frac{\mathrm{d}\mkern 0.0mu\,r_{EH}}{\mathrm{d}\mkern 0.0mu\eta}}} \displaystyle=-1, (39) see definitions in equations ( 13 )-( 15 ). Consequently, the presence of an accelerating/decelerating regime directly influences the (conformal) growth rate of the AH, but not that of the PH and EH. In adition, r_{AH}(t_{s})=r_{AH}^{\star}<\infty in the FTPR model ( 16 ), whereas it diverges for the SFS ( 1 ); see Fig. 4 . V Raychaudhuri averaging and related strength of exotic singularities The general SFS scenario and the FTPR found here are weak singularities Nojiri et al. 2005 according to the definitions of Tipler Tipler 1977 and Królak Królak 1986 , and therefore extendible Fernández-Jambrina and Lazkoz 2004 . A different definition of singularity criterion was given by Raychaudhuri Raychaudhuri 1998 , and later developed by one of the Authors Da̧browski 2011 . According to these considerations, one is always able to take an average of any physical or kinematic scalar quantity \chi over the entire (open) spacetime in the form <\chi>\equiv\left[\frac{\int_{-x_{0}}^{x_{0}}\ldots\int_{-x_{3}}^{x_{3}}\chi\sqrt{-g}d^{4}x}{\int_{-x_{0}}^{x_{0}}\ldots\int_{-x_{3}}^{x_{3}}\sqrt{-g}d^{4}x}\right]_{\lim_{x_{0},\ldots x_{3}\to\infty}}. (40) For a FLRW cosmological model, we focus on averaging the acceleration scalar \chi=\theta_{,\mu}u^{\mu}=3H^{2}(q+1) , where \theta is the expansion scalar, q=-\ddot{a}a/\dot{a}^{2} is the deceleration parameter, \mu,\nu=0,1,2,3 . In our case, instead of infinite limit of the coordinates, we will limit ourselves to averaging on the domain of applicability of the model. Then, according to ( 40 ) Da̧browski 2011 <\dot{\theta}>=\lim_{\genfrac{}{}{0.0pt}{}{t_0 \to 0}{t_1 \to t_s}}\frac{3\int_{t_{0}}^{t_{1}}a^{3}\left(\frac{\ddot{a}}{a}-\frac{\dot{a}^{2}}{a^{2}}\right)dt}{\int_{t_{0}}^{t_{1}}a^{3}dt}. (41) For our decelerating model ( 1 ), on the one hand, the denominator of ( 41 ) gives \displaystyle\int_{0}^{t_{s}}a^{3}dt=\int_{0}^{a_{s}}\frac{a^{3}da}{\dot{a}}=\frac{1}{H_{\star}a_{s}^{n}}\int_{0}^{a_{s}}\frac{a^{4}da}{\left(1-\frac{a}{a_{s}}\right)^{n}} (42) \displaystyle=\frac{a_{s}^{5-n}}{H_{\star}}\left[\frac{1}{1-n}-\frac{4}{2-n}+\frac{6}{3-n}-\frac{4}{4-n}+\frac{5}{5-n}\right], which is finite. The numerator of ( 41 ), on the other hand, gives \displaystyle 3 \displaystyle\int_{0}^{t_{s}}a^{3}\left(\frac{\ddot{a}}{a}-\frac{\dot{a}^{2}}{a^{2}}\right)dt=3\int_{0}^{a_{s}}a^{3}\left(-\frac{1}{\dot{a}}\frac{\ddot{a}}{a}+\frac{\dot{a}}{a^{2}}\right)da (43) \displaystyle= \displaystyle-3H_{\star}a_{s}^{n}\int_{0}^{a_{s}}n\left(1-\frac{a}{a_{s}}\right)^{n-1}da \displaystyle- \displaystyle 3H_{\star}a_{s}^{n}\int_{0}^{a_{s}}(2-n)\left(1-\frac{a}{a_{s}}\right)^{n}da=-\frac{9H_{\star}a_{s}^{n+1}}{n+1}, by using both ( 1 ) and ( 3 ). Clearly this contribution is also finite. Therefore, the SFS which appears in the decelerating model ( 1 ) is weak in the sense of Raychaudhuri spacetime averaging (cf. Da̧browski 2011 ). For the accelerating model ( 16 ), note that the denominator in the Raychaudhuri averaging ( 41 ) is a finite and positive quantity as the co-moving 3-volume is a smooth positive-defined function on that compact integration range. For the numerator, expressions ( 16 ) and ( 17 ) lead to \displaystyle 3 \displaystyle\int_{0}^{t_{s}}a^{3}\left(\frac{\ddot{a}}{a}-\frac{\dot{a}^{2}}{a^{2}}\right)dt (44) \displaystyle= \displaystyle 3\int_{0}^{a_{s}}\left(-2H_{1}+H_{2}a_{s}^{n-1}a^{3}\left(1-\frac{a}{a_{s}}\right)^{n-1}\right)da \displaystyle= \displaystyle-6a_{s}H_{1}+3a_{s}^{n+3}H_{2}\left(1-\frac{3n}{n+1}+\frac{3n}{n+2}-\frac{n}{n+3}\right), which is again finite. This shows that the FTPR singularity which appears in the accelerating model ( 16 ) is also weak in the sense of Raychaudhuri spacetime averaging. VI Conclusion In this paper, we have proposed a method to construct solutions of the cosmological field equations by prescribing a specific functional form of the Hubble parameter capable of inducing different types of future cosmological singularities. This approach provides a particularly straightforward framework for analysing the ECs within general relativity, which in turn allows one to characterise the nature and strength of the resulting singularities in relation to the notion of geodesic incompleteness. Within this approach, we have explicitly constructed two new models driving the cosmological expansion towards a standard SFS and a novel FTPR (finite-time pseudo-rip) singularity. Both singularities exhibit a similar structural behaviour in that the cosmic time of occurrence, the scale factor, and the (effective) energy density remain finite, while the pressure diverges, i.e. |p|\to\infty as t\to t_{s} . On this basis, one might be tempted to classify both events as SFS Barrow 2004a (or Type II singularities in the sense of Ref. Nojiri et al. 2005 ). However, we have argued that such a classification is misleading, since the two scenarios exhibit fundamentally different behaviour when analysed from the perspective of the ECs. While the original SFS was introduced as a scenario in which only the DEC is violated Barrow 2004a ; Barrow 2004b , the model ( 16 ) instead, displays a phantom-dominated future evolution that leads to the violation of all ECs. Although only two example models have been addressed, the method can be applied to generate more solutions containing a rich set of exotic singularities in cosmology. Some examples of these singular behaviours that have attracted the attention of many authors in recent years (for a review see Refs. Nojiri et al. 2005 ; Da̧browski 2014 ), but for our study the most appropriate were a big-rip, a little-rip, a little sibling of the big-rip, a pseudo-rip, and a big-freeze. We have also presented some systematics to study the energy conditions in general relativity which allow to characterise the types of singularities in respect to their strength related to the notion of geodesic incompleteness. In fact, both singularities considered in this work are of a weak type and the evolution of the universe can be extended beyond them. We have also discussed the global characteristics of the models at hands by using the method of Penrose diagrams. In particular, we have investigated the structure of cosmological horizons (event, particle, and apparent), as well as the asymptotic structure of singularities. Though only a limited time of evolution of the universe between the standard BB radiation-dominated singularity at time t=0 to the future exotic singularity at time t_{s} has been addressed, due to the fact of geodesic completeness of exotics, one may extend them symmetrically into a future region evolving towards another big-crunch singularity as in Ref. Da̧browski and Marosek 2018 ; thus, making kind of a cyclic universe modelling. Finally, we have found by using the method of the so-called Raychaudhuri averaging that the two models under study are also weak with respect to this definition. We hope both the method of generating the weak singularity solutions and the method of Raychaudhuri averaging may prove successful tool in future cosmological investigations including their observational verification. Acknowledgements. T. B. V. acknowledge the hospitality of the University of Szczecin where this work was partially conducted. He also acknowledges financial support from Universidad Complutense de Madrid and Banco de Santander through Grant No. CT63/19-CT64/19 during that period.