Study of Device for Hanging and Mounting AOD Retort
80-tonne steel structure subjected to high thermal and mechanical stresses, had an obsolete fastening system that no longer met the current ergonomic requirements. The study of a new fastening system was based on the specifications requiring significant ergonomic improvement, a resistance to certified stresses, limited space requirements and cost, as well as ease of implementation and operation.
The retort consists of a refractory steel envelope, supporting strong temperature gradients, and numerous layers of refractory materials, providing thermal insulation between the molten pool and the envelope.
The retort is fastened to a belt by six specific screws, which ensure permanent plating in all directions required by the process steps.
These screws, which are very massive (M64), must be inserted inside the belt through small loopholes, and then lifted at arm's length to complete the assembly. The regular change of retorts for the replacement of refractory materials makes this task difficult.
Therefore, the project includes a study of an alternative fastening that no longer requires the insertion and manual maintenance of heavy axles in the belt. After analyzing the customer's proposal, a solution using tilting axes outside the belt was proposed and adopted. While the ergonomic gain is significant (the accessibility to the elements is strongly increased and the efforts to be provided strongly decreased), the mechanical stresses are increased by the distance between the fasteners.
The entire existing retort was thus modeled in its original version, and the necessary modifications were made: adding plates, reinforcements, gussets, axes, trimming of colliding elements, and positioning peripheral accessories.
Once the 3D model was completed, a complete recasting in the median surfaces for the sheets and in external surfaces for the solids was carried out within the assembly, with respect of an export in a *sat file for the integration in RFEM 5.
Integration into RFEM 5
This first import failed due to the precise modeling of the refractory layers, which generated a large number of surfaces and lines, that were either superimposed, offset, or absent. After several unsuccessful attempts to clean up the model, it was decided to make a new export from a 3D sketch that only contains the edges of the CAD model. This implied a manual, delicate, and tedious reconstruction of all surfaces under RFEM, using quadrangle models, and other polygons and surfaces defined by contour. The modeling was very complex, because the retort has cone-like surfaces or cylindrical components with reliefs, and some elements have openings of complex geometry on non-planar surfaces.
This step was necessary to model the distribution of the masses of refractory and molten steel within the retort as accurately as possible, for all the key directions. Since the dead load is the only stress applied to the model, it was essential that the geometries respect the real situation as closely as possible.
The second complexity was to model the interaction of the retort on the belt by simulating a sliding contact with a tight bond and stops. It took several iterations to find a stable model working in a non-linear mode, which means more and longer iterative calculations than in linear approach.
The results obtained were used to verify the existing assembly on the one hand, and the proposed solution on the other hand.
Do you have further questions or need advice? Contact us via phone, email, chat, or forum, or search the FAQ page, available 24/7.
For plastic stresses of ductile materials such as steel, yielding effects occur where the increase in deformation comes along with hardening. They lead to permanent deformations - and in extreme cases, to the destruction of the structural component.
- I want to create a mapped mesh for a circular hole plate. Is it possible to generate such a mesh in RFEM?
- Is it possible to set user-defined values when viewing solid stress results?
- How can I create a drilled beam in RFEM?
- How can I additionally model welds for the connection of two surfaces with a contact solid?
- I would like to use surface releases to model a bolt connection that can only absorb compressive stresses from the support. Unfortunately, I cannot do this because my model is always unstable. Is there a template model?
- How can I display shear stresses on null elements in SHAPE‑THIN?
- I have calculated a box girder. Which surface results or surface stresses can I use to evaluate the buckling behavior of the web plates?
- When creating a material, there are no nonlinear material models available for surfaces or solids. Why?
- What is the formula used to calculate the ASCE 7 wind velocity profile?
- What is the difference between the Nonlinear Elastic and Nonlinear Plastic 2D/3D material model?