Spring coils are defined as spring-open or fall-down coils that occur after the coiling, evacuation, transport, storage, or further processing steps of hot rolled steel. These coils pose a significant safety risk, and the transportation of these problematic coils away from the strip mill becomes more difficult. In addition, the productivity of the entire downstream path can be affected by spring coils. The parameters with the highest impact on the coil stability are strip thickness, coil weight or diameter, resistance to deformation or yield stress under specific coiling conditions, temperature-dependent Young ¿s modulus, geometrical conditions of bottom rolls of coil car and the friction coefficient between the layers and external points of contact from the rolls. The framework of this thesis includes a finite element (FE) simulation model of the coiling process in Abaqus, explaining the general structure and approach to analyse and predict the spring effect. This simulation model provides initial results by varying the strip thickness and material properties. In general, the tendency of the spring effect decreases when the strip gauge or yield stress decreases. This basic model was extended to best mirror the process conditions. After successfully implementing the modifications, the parameter variation continued in order to gain a better understanding of the spring effect and its influencing parameters. Throughout this parameter variation, the strip tension, strip length, strip end position and bottom roll distance were altered. An example of the outcome was that at transition region (no spring effect vs spring effect) higher strip tension tends to stabilize the coiling process and thus no spring effect is observed. While the average computing time for simulating a coil with, for example, a 45 m long and 16 mm thick strip takes more than 10 hours, research has been conducted on a new approach. The introduction of an Abaqus user subroutine called SIGINI in the simulation model has been presented. Due to this subroutine, the simulation results are obtained faster. In this example, it takes less than an hour to obtain the results. Moreover, a stress comparison was performed between the basic simulation model and the user subroutine model. The stress values showed a good correlation, and as a rule, a maximum stress deviation range of ± 14 % is expected for the entire coil from the subroutine simulation compared to the basic elastic simulation. The mean Mises stress deviation comes to 5.38 % for all arbitrarily chosen reference nodes of the coils.