Collaborative work on aircraft wing design in pSeven Enterprise
Industry: Aerospace| Product: pSeven Enterprise
June 10, 2025
Introduction
Designing an aircraft wing is a complex challenge. The behaviour of the wing in the flow is highly nonlinear and obtaining valid wing properties through simulation requires significant experience from an aerodynamics engineer. However, the design process does not conclude with a single iteration; it involves evaluating trade-offs between multiple aerodynamic parameters. Traditionally, selecting the best design was a trial-and-error approach, heavily dependent on the designer’s intuition and experience. Today, with the ability to explore a vast range of design options, numerical optimization has become the preferred method.
pSeven Enterprise provides the tools to perform such optimization studies in an automated manner, along with additional features that enhance overall efficiency of the process:
- Sometimes it is necessary to repeat the optimization study, for example for a new flight regime or when other aspects of the problem change. With pSeven Enterprise, this process is straightforward, as the workflow can be easily restarted with new inputs once the process has been captured.
- pSeven Enterprise helps minimize the risk of human error by automating data flow between blocks, which reduces the chances of unit conversion mistakes or copy-paste errors. Any result, even if not preserved, can be later reproduced with exact same process and settings.
- Workflows in pSeven Enterprise can be easily shared with other users to delegate tasks to colleagues. For instance, a simulation expert can assemble the workflow, allowing an intern to use it to generate the same results with a fixed methodology, protected from changes in critical aspects. Such approach enables capturing engineering methodologies and preserving knowledge within the organization.
- Design optimization problems are typically addressed by a team of engineers rather than a single individual. pSeven Enterprise offers a collaborative environment that enables team members to share materials and work together to assemble multidisciplinary workflows with efficient division of responsibilities.
While creating a workflow in pSeven Enterprise, especially involving several disciplines and departments, it is necessary to define a format to describe the input and output data. In the case of wing optimization, this format is used to represent both the initial and optimal wing geometries and reflects the exchange format definition, which typically already exists in some form. It can take various forms, such as a CAD model, a text file with a custom description or a set of geometric parameters. Once this format is established, it can be used as a basis for setting up collaborative work, as all participants in the process will need to adhere to this standard. In a simple case, an object prepared by one department can be treated as an input by another.
Usually, data transfer between departments during multidisciplinary analysis is organized manually. This leads to the following problems:
- Collaborating across departments requires effective communication channels and project management. Poor coordination can result in delays and confusion.
- Merging simulation results from various departments (e.g., structural, fluid dynamics, electrical) is a complex process. Each department may generate data in different formats, units or scales, complicating integration.
- Departments may have different timelines for completing their simulations, which can create bottlenecks if one department is waiting for input from another.
- Even if the processes in each department are ideal, assembly of collaborative fully automated simulation chain may be tricky.
In pSeven Enterprise, multidisciplinary analysis can be organized as a workflow that references other individual study workflows, which are created and maintained by different teams within the company (Fig. 1). By referencing these individual workflows, the team gains the flexibility to assemble a complete workflow while still allowing each study to be edited, run independently and reused across multiple workflows. In this case, data exchange is automated and does not require a human in the loop.
Figure 1. Workflow referencing
In addition, to enable knowledge reuse between departments and preserve competence, custom user blocks can be developed. These blocks provide a no-code or low-code approach to integrate external simulation tools or implement proprietary design methods. They can help to significantly reduce routine operations effort due to tailored interfaces, capturing only essential settings, fixing internal complexity and even protecting it from unauthorized altering.
Problem statement
This article describes how the features presented above can be applied to real-world examples of wing design.
The first challenge addresses the iterative interaction between aerodynamic and structural analyses. The wing is considered deformable, meaning that the force factors obtained from aerodynamic simulations can alter the wing geometry. Structural analysis is then used to estimate these deformations. Due to the changes in the geometry, aerodynamic analysis must be performed again to correctly determine the forces acting on the wing. Such balancing process is repeated iteratively until the changes between steps become negligible. The result of this study is the true shape of the wing and the refined loads acting on it.
The second challenge focuses on optimizing the wing geometry to improve its aerodynamic characteristics. The goal is to minimize drag force while maintaining lift force by adjusting the twist angles of the wing cross-sections.
Since these problems are related, they can share the same approach for geometry description, model preparation and other setup steps.
Wing parametrization
To characterise a single cross-section (Fig. 2), parameters such as profile type, leading edge coordinates, chord length, rotation axis coordinates and rotation angle can be used.
Figure 2. Wing cross section
It is then possible to describe the wing as a 3D surface specifying the parameters of all cross-sections that form the wing. This information can be saved in a single text file (Fig. 3), which will later be used as both input and output for the workflows.
Figure 3. Wing description with a text file
Iterative search
The first problem addressed in this article involves a multidisciplinary analysis to determine the true deformed shape of the wing. This methodology is based on finding the equilibrium point between the external aerodynamic moment generated by the incoming airflow and the internal structural moment of the wing. First, the aerodynamic analysis estimates the force factors acting on the wing, which are then applied as loads in the structural simulation to calculate the resulting deformation. Because the shape has changed, the aerodynamic analysis is repeated to obtain a new set of force factors. This computational cycle (Fig. 4) continues until the changes in geometry become negligible.
For simplicity, only the pitching moment is considered in this article, reducing the structural problem to pure torsion. However, this approach can be extended to any set of force factors. As a result, the changes in geometry include only modifications to the twist angles of the wing cross-sections.
Figure 4. Computation cycle
A key aspect of this process is the involvement of different types of simulations. Let us consider a separate workflow for each discipline.
The workflow for aerodynamic analysis (Fig. 5) consists of a single block with AeroSandbox integration. It receives arrays of cross-section parameters describing the geometry and produces a list of moments calculated at each section. Normally, AeroSandbox can be integrated using a Python block and a custom script, but since most calculations in this tool are easy to formalize, the aerodynamic engineer developed a user block that includes all solver settings within its ports. Each type of simulation is assigned to a specific preset of the block. This reusable block can be used in other workflows that require AeroSandbox, benefiting both the block’s developer – who saves time during integration setup – and other users who may not be familiar with the software integration process.
Figure 5. Aerodynamic analysis
The workflow for structural analysis (Fig. 6) is more complex. It includes the preparation of input files via the “Prepare study” block, simulation in Salome-Meca and post-processing using the “Extract angles” block. Salome-Meca is integrated using a “Program SSH Linux” block, which serves as a connector to a remote Linux machine and executes the task there. The inputs to this workflow are geometry parameters and moments, while the outputs are the twist angles of each cross-section.
Figure 6. Structural analysis
The complete workflow for solving the problem is presented in Figure 7. First, two Text blocks are used to parse the input files: one for the wing geometry description and another for the stiffness line description. Next, a Python block identifies the intersections between the cross-sections and the stiffness line; the resulting points later serve as the rotation axis. Once these preparations are complete, the simulation cycle begins. The cycle is controlled by a “While” block, which checks whether the termination criteria are met and triggers the next iteration.
The simulation process is located within the “Simulations” composite block, where each single domain workflow is imported using the workflow reference feature, ensuring that each run of the main workflow incorporates the most up-to-date version of the child workflow, including any recent changes. The twist angles obtained in the current iteration are compared to those from the previous one using a Python block. Once the convergence criteria are met, the final design and moment values are sent to the output ports. The workflow is constructed in such a way that, to set up a new run, the user only needs to provide new input files, while the expected results will be available at the output ports.
Figure 7. Deformed shape search workflowe
Optimization problem
Figure 8. Aircraft force factors
The optimization problem in this article is formulated as follows: to obtain a design variant in which the drag force is minimized while maintaining the initial lift force. The control parameters in this case are the rotation angles of the individual cross-sections:
\[ \begin{cases} C_x \to \min \\ C_y \geq C_y^{\text{original}} \\ \left\{ \begin{array}{l} a \leq \delta \theta_1 \leq b \\ a \leq \delta \theta_n \leq b \end{array} \right. \end{cases} \]