how to adjust the Ca/Si and Al/Si ratio of CSH during process simulation of blended-cement hydration

[dangerous-pocket]

Hello everyone,

I am currently using GEM-Selektor to perform process simulations of cement–fly ash blended systems, and I would like to ask for some guidance regarding the control of the chemical composition of C–S–H gel, specifically the Ca/Si and Al/Si ratios.

From experimental studies, it is well established that compared with pure Portland cement systems, the incorporation of supplementary cementitious materials such as fly ash generally leads to:

a decrease in the Ca/Si ratio of C–S–H, and
an increase in the Al/Si ratio, due to Al incorporation into the C–S–H structure.

However, I am not fully clear on how to properly reproduce or control this compositional evolution of C–S–H gel within a process simulation in GEM-Selektor.

In particular, I would like to understand:
How to control the Ca/Si ratio of C–S–H in blended systems compared with pure Portland cement, in agreement with experimental observations.
How to enable and control Al incorporation into C–S–H, and whether it is possible to impose or constrain a target Al/Si ratio within the C–S–H (or C–A–S–H) phase.

Any advice, references, or example setups would be greatly appreciated.
Thank you very much for your time and help.

Best regards,

[dangerous-pocket]

the Cemdata18 dataset was used durning calculation

[dangerous-pocket]

If anyone happens to know the answer, I would greatly appreciate a reply, as it would be of significant help to me. Thank you very much.

[gdmiron]

In a GEMS thermodynamic equilibrium calculation, the user specifies the bulk cement composition as input, and GEMS computes the resulting phase assemblage and hydrate compositions based on the thermodynamic properties of all species and phases in the system (e.g., from cemdata18). Consequently, the Ca/Si and Al/Si ratios in C–S–H are outputs of the calculation. They are determined by the bulk Ca–Si–Al composition of the system and by the thermodynamic description of the CNASH solid-solution model.

Directly prescribing the CNASH composition is therefore not possible. Instead, the CNASH composition can only be influenced indirectly by:

modifying the bulk input composition, or

modifying the thermodynamic properties of the CNASH end members.

The CNASH end members are parameterized based on experimental data, but they may not perfectly represent all modeled systems. For example, applying a positive correction to the standard Gibbs energy (G⁰) of a given end member decreases its thermodynamic stability, thereby reducing its amount in the equilibrium assemblage and altering the overall CNASH composition. Any such modification, however, must be justified.

An alternative way to constrain the CNASH composition is by adjusting the lower and upper KC bounds, which define the minimum and maximum allowable amounts of individual end members. By assigning different KC limits to specific end members, it is sometimes possible to obtain the desired phase composition and elemental ratios.

A recommended workflow is to first perform an unconstrained calculation and then gradually change the KC bounds around the calculated values until the target composition is reached. This approach helps avoid numerical instabilities or wrong solutions.