Hi uno.
A. Please forget my first message.
B. Sorry for this delayed response. My attempt to understand the differences between the three currents took me more than 1-2 days that I expected. I have been checking for and thinking to the three currents over and over again. I have come to some conclusions, but, as I am not 100% sure, I recommend you to take my below notes as suggestion only.
1. The name of Icw is "rated short-time withstand current" and has nothing to do with breaking capacity (this confused me). The physical meaning of Icw is more close to the making current Icm than to breaking current(s).
2. The making current Icm must be biger than the peak value ip of shc current and specifies the maximum curent which the CB withstand to in respect to the electro-dynamic forces produced by the inrush current of shc.
3. For CBs with intentionally time delay and resp. with Icw, the CB must be able to withstand to the shc current (el-dynamic forces) for the duration of the set delay (e.g. 100 msec) and then to break/trip the circuit (Icu) if the shc is still present. This is why I afirm that the Icw is more close in meaning to the Icm. The Icw is to be bigger than r.m.s. value of iksym assumed to be constant from the begining of shc (i.e. 40 kA for 1/3 seconds in your case).
4. Utilization category A means that the CB is not intended to be used in series with another protection device in supply line; this would be coresp. to the motor protection CBs of Siemens.
5. Utilization category B means that the CB is intended to be used in series with another protective device in supply line; this would be coresp. to the line protection CBs of Siemens. Consequently, in order to ensure the selectivity of protection, the CB is to be provided with and intentionally delay release time (which can be adjusted) and, hence, with the Icw.
6. Based on notes 2 and 3 it results that the Icw is only defined/specified for CBs used in category B.
7. The difference between Icu (ultimate) and Ics (service) currents consists in the qualitative and quantitative conditions established for testing the CBs. I cannot tell exactly what is this diference.
8. The CBs are constructed in four cases, depending of relationships between the three currents (equal, >, <, <>), i.e. Ics=Icw<>Ics, Ics=Icw=Icu etc.
9. I have checked the Siemens products catalogue and I haven't found specified the Icw value. I assume that all Siemens CBs are constructed with Ics=Icw. On the other hand, the Siemens offers CBs provided to "azn" electronic device; some CBs are only with "an" device - coresp. to motor protection while others are provided with "azn" device. The Icw and its delay time can be adjusted on "z" portion of tripping characteristic, resp. from Icw=0...8*In of CB and td=0...300 msec; t=0 msec coresp. to motor protection CB; t>0 sec coresp. to line protection CB (categ. B); the CB can be used for both categories.
10. The Siemens states that back-up fuses are only required if the shc current at the point of installation exceeds the Icu (ultimate) value of CB. Cumulated with the fact that Ics is in almost cases smaller or equal with Icu, it result that the base for selection the breaking capacity of CB is the Icu, but as I said above, I cannot tell for sure what is the difference between Ics and Icu. A manufacturing engineer may tell this difference; I am only an installation engineer.
11. The breaking capacity/curent of CB (Icu or Ics) is to be bigger that the shc current evaluated at the time when the CB's contacts start to separate. This value is to be based on the r.m.s. value of the iksym at the time when the contacts start to separate; I personally prefer to calculate the instantaneous value of shc current (ikasym+iksym) at the respective moment.
12. As far as I am concerned, I "split" the short-circuit process into three sequences:
a) t=0...topening: the CB must be able to withstand to the el-dynamic forces produced by the inrush current of shc; therefore, Icm > ip (please read the ">" sign as "must be bigger than"

; for CB of category B this additional condition must be met: Icw > iksym(t=0) - acc. to note 3.
b) t=topening: the CB must be able to break/trip the shc current at this time; therefore, Icu > Ik(r.m.s. of iksym, t=0 (assumed constant)) or, in my theory, Icu > ik(t=topening).
c) t=0...tend: the CB must be able to withstand to the heat effects produced by the shc current in the interval t=0...topening and by electric arc between opened contacts in the interval t=topening...tend; therefore, Ilt > Ith (thermal current).
I hope my notes would be of a help to you, but as I said before, I recommend you take this scheme as suggestion only.