Discussion forum for Serpent users

What exactly do you mean by multi group cross sections? The internal routines are based on one energy group. The number of groups makes no difference for the depletion solver since the reaction rates are integrated over all energies.

You have to define the power density using "set powdens" card. If you have the power density fixed, then the problem is most likely related to the material volumes. Serpent cannot calculate the pin volumes in the second case, so the values are initialized to unity. This changes the normalization com...

Unfortunately, no.

How do you normalize the results? Do you define the total power or power density?

Using a cell detector (dc option in the cell card) is one way to do it. Serpent doesn't divide the result with volume, so you have to do this manually, or by setting the dv option. If you are looking at dose rates in discrete locations, I recommend you use the super-imposed track-length detector ins...

This warning is related to the calculation of CMM diffusion coefficients. The flux is zero in one of the energy groups (this can happen especially in fast systems). The message will be changed into something more sensible in the next update.

It looks like the problem is related to the energy boundaries of unresolved resonance probability table sampling, and one unfortunate subroutine that tests them even if the sampling is not on. Try changing the if-statement on line 75 of dopmicroxs.c from: if((RDB[rea + REACTION_XS_I0] > 0) || (E > R...

Could you send me your input that produces this error? I've worked on the thermal scattering data since July and I could check if this problem has been fixed with the modifications.

In your model universe 102 is set as the background universe in the pbed card (the universe that fills the space between the pebbles), but also as the universe that fills each of the two pebbles (in file pbed.inp). However, this universe is not defined anywhere in the input. What you need is: 1) A u...

There are no limitations when it comes to parallelization, but if the model is very large, you may eventually run into problems with large memory footprint. One limitation related to group constant generation is that the method works on universes, so the homogenized region has to be represented as a...