KnowMesh
← Back to search

Marta Michalak

Loads Engineer · Flight Dynamics & Aeroelasticity

Warsaw, Poland

Loads Engineer responsible for generating the design load conditions used for aircraft structural sizing and certification. Covers the full load cases spectrum from symmetric and asymmetric maneuvers to ground loads, and interfaces closely with structural analysis and aeroelasticity teams to ensure loads are correctly applied to FEM models. Builds Python automation for loads envelope generation and loads-to-FEM translation.

Expertise

  • maneuver loads envelope computation
  • ground loads (taxi, landing, tow-bar, jack)
  • limit and ultimate load condition generation
  • loads redistribution with structural flexibility
  • loads-structure interface for FEM sizing

Technologies

NASTRAN Python MATLAB MSC Patran pandas Linux HPC Git

Work History

2025-01

Tow-bar and ground handling loads — computed limit loads for aircraft ground handling (towing, pushback, parking on slope) per CS-25.509. Identified critical tow-bar attachment load that drove nose gear leg sizing.

Challenge: The CS-25.509 tow load case at maximum ramp weight on a 15° slope produced a lateral tow-bar load that was not covered by the existing nose gear FEM model (assumed symmetric loads only). Required adding lateral load path modeling to the nose gear FEM — a late-stage FEM scope addition.

Learned: Ground handling load cases must be assessed against FEM model scope early in the structural design phase. CS-25.509 lateral tow loads are a standard requirement but are frequently overlooked until the loads-to-structure interface is made — discovering an FEM scope gap at that stage is costly.

MATLAB Python NASTRAN

2024-06

Loads-structure interface — developed Python tool to convert aeroelastic loads from NASTRAN cut-sections to applied nodal loads on the structural FEM for stress analysis. Handled load redistribution across multi-path structure.

Challenge: Load redistribution across the multi-path wing box (skin, spar caps, ribs, fasteners) required interpolation of NASTRAN force/moment resultants to nodal loads — directly applying section resultants as nodal forces introduced artificial stress concentrations. Required moment-balanced nodal load application using constraint equations.

Learned: Loads-to-FEM transfer must preserve moment equilibrium at every structural section. Naive nodal force interpolation without moment balance artificially stiffens the structure at load application points. Use constraint equations (MPC) to distribute section moments correctly to the surrounding structural nodes.

Python NASTRAN pandas

2023-11

Landing loads analysis — computed limit loads for normal landing, hard landing, and tail-down landing conditions per CS-25.473 and CS-25.481. Provided loads to landing gear structural team and wing carry-through structure.

Challenge: Hard landing at maximum sink rate (3.05 m/s) produced dynamic landing gear reaction loads that were 40% higher than quasi-static analysis. Required a multi-body dynamic simulation of the landing gear oleo-pneumatic compression to capture the dynamic load amplification correctly.

Learned: Landing loads for hard-landing conditions cannot be computed with quasi-static analysis alone — the dynamic amplification from the oleo-pneumatic response is significant and underestimated by static methods. Multi-body dynamics simulation is required for accurate hard landing structural loads.

MATLAB Python NASTRAN

2023-05

Symmetric maneuver loads envelope — generated critical flight loads for 1,200 symmetric maneuver points (altitude, Mach, n-factor combinations) per CS-25.331. Extracted critical conditions for wing bending, shear, and torsion sizing.

Challenge: Generating 1,200 load cases with aeroelastic correction was computationally intensive — sequential processing took 3 weeks. Parallelized the NASTRAN runs across 20 HPC nodes using a Python job manager, reducing wall time to 4 days.

Learned: Loads envelope computation at CS-25 scale requires systematic HPC job management. Building a Python-based NASTRAN job manager (submit, monitor, retrieve, post-process) from the start of the loads campaign is a week-one investment that pays back 5x over the full loads cycle.

NASTRAN Python MSC Patran Linux HPC