The issue of force calculation when setting a moving part

Hello marton,
I want to simulate the molecular impact force on the windward surface of a spacecraft in orbit. Below is my model.

The small rectangular box represents the spacecraft, and the larger one represents space.

Theoretically, both configurations—setting the surrounding space in motion while keeping the spacecraft stationary or setting the spacecraft in motion while keeping the surrounding space stationary—should achieve my goal.

However, the force obtained from these two setups differs significantly, and the results from the second setup are clearly inconsistent with reality.

  1. Surrounding space is moving part, spacecraft stationary
    As shown in the figure, I set vz = 7800 m/s and defined space as a moving part .

The force on the windward surface (left side of the small rectangular box) is 1.045e-4 N.

The pressure distribution in space is as follows, which aligns with my intuitive expectations:

  1. Spacecraft is moving part, surrounding space stationary
    As shown in the figure, I set vz = -7800 m/s and defined the spacecraft as a moving part .

In this case, the force on the windward surface is 4.5e-5 N. The pressure distribution in space is shown in the figure. Surprisingly, there is pressure inside the small rectangular space, which seems unreasonable.

Even more problematic, when I tested by setting vz = -78000 m/s (spacecraft as moving part, surrounding space stationary), the forces on the windward and leeward surfaces were almost identical!

image

Theoretically, I should set the spacecraft in motion while keeping the surrounding space stationary to achieve my goal. However, it seems that defining the moving part in Molflow this way leads to issues.

My Questions:

  1. Can I achieve my goal by setting the surrounding space be moving part while keeping the spacecraft stationary? Is the result of this setting reliable? Why?
  2. Is defining the spacecraft as a moving part currently unfeasible, and does it produce unreasonable results?

Attached is my model.
test.zip (253.8 KB)

Dear Wang,

I’ve written the documentation for the moving parts feature, and created a cubesat example similar to yours. I tried to answer your questions in it, but let me know if not everything’s clear. Thanks for bringing up the topic.

Cheers, Marton

Thanks, Marton!

The document is very detailed and has resolved my previous questions. However, one issue remains—how can I validate the simulation results for the moving part? Are there any papers or trusted results that can be reproduced using MolFlow?

Hello Wang,

We collect Molflow-published articles under the Gallery/Results → Articles section of the website.

In the 2024 section, there is an article that used Molflow to measure the capture coefficient:

Numerical Studies of Ram-Air Intake for near Earth Satellites
Nishita Ravuri, Ashish Vashishtha and Stephen Scully
AIAA SCITECH 2024 Forum, 8-12 January 2024, Orlando, FL
https://doi.org/10.2514/6.2024-2862

I think this might be relevant.
It is possible that @rkerseva might be aware of other papers.

Cheers, Marton