Pressurization - Discrepancies in block behaviors as part of a pressure boundary
I was building a test case for a large blast door which would become part of a pressurization boundary. This is due to the 4 block width maximum opening in the direction of slide of the hangar door blocks.
I noticed some odd behavior in how it acted as a pressurization boundary. I therefore set up a test and compared the behaviors of a variety of blocks. I presumed that this game simulated the basic physics of gas containment and that a gap in the pressure boundary would leak. I set up the test cell so a bright light is used to find gaps.
I was surprised to find that blocks that completely passed the light test leaked and blocks that utterly failed the light test sealed (were leak tight). Some of the observed behaviors make no sense.
See attached test world file. Here is link to video of the discrepancies found (https://youtu.be/_hcfoRgfTP0).
I recommend the following:
- Choose a philosophy for determining pressure boundary behavior. I recommend the light test for gaps.
- Based on what I found some minimally large gaps will be deemed acceptable
- Review all blocks for this behavior.
- Fix the ones that don't fit the philosophy.
At a minimum the following need to be fixed:
- Blast door edge parts (orientation matters as these are 1x1x0.9)
- Small oxygen tank
- Door and Offset door
- Round column
- Small warfare reactor
- The sealing behavior of a slab of blocks (e.g. blast door) blocking an opening in order to provide a pressure boundary. See end of test video for this case.
Regards,
Bill
I have the same bug
Hello, Bill,
thanks for letting us know and for the commented video!
I did watch it, tried it on my end as well and most of the mentioned blocks are indeed behaving at least inconsistently, if not wrongly all along.
But for some other blocks, I think that those are fine. Namely: Cylindrical column and Small warfare reactor - even trying with you light test, there are visible holes where the models are not touching the wall if you take a closer look. So I did not add these to the bug.
Also did not wrote down the case from the end of the video, as I believe that is right as well. Yes, you are right it "looks" like the blast door sitting on the window are touching it without any holes, but... there is still vacuum for sure in between the window block and the blast door block. Plus the fact that they are sitting on each other does not mean that they are (properly) connected. You can even see (by the colors in control panl) that the pistons and hinges are considered different grid. Hope you understand.
However, in all other blocks you were right and I did put them into our internal system as a bug.
Kind Regards
Keen Software House: QA Department
Hello, Bill,
thanks for letting us know and for the commented video!
I did watch it, tried it on my end as well and most of the mentioned blocks are indeed behaving at least inconsistently, if not wrongly all along.
But for some other blocks, I think that those are fine. Namely: Cylindrical column and Small warfare reactor - even trying with you light test, there are visible holes where the models are not touching the wall if you take a closer look. So I did not add these to the bug.
Also did not wrote down the case from the end of the video, as I believe that is right as well. Yes, you are right it "looks" like the blast door sitting on the window are touching it without any holes, but... there is still vacuum for sure in between the window block and the blast door block. Plus the fact that they are sitting on each other does not mean that they are (properly) connected. You can even see (by the colors in control panl) that the pistons and hinges are considered different grid. Hope you understand.
However, in all other blocks you were right and I did put them into our internal system as a bug.
Kind Regards
Keen Software House: QA Department
I agree the last case is an odd one. I take it that the "different grid" characteristic is significant. Using real world physics, when the hinged door closes, it "seals" to the blocks surrounding the windows. It is leak tight based on the light test. I agree there might be a small volume between the window glass and the blast door blocks that would be under vacuum. However when the window is removed the small volume should almost instantly pressurize. The simulation evidently can't determine a seal or assumes a separate grid cannot be part of a seal. This is too bad because it limits the size of a hangar door.
Regards,
Bill
I agree the last case is an odd one. I take it that the "different grid" characteristic is significant. Using real world physics, when the hinged door closes, it "seals" to the blocks surrounding the windows. It is leak tight based on the light test. I agree there might be a small volume between the window glass and the blast door blocks that would be under vacuum. However when the window is removed the small volume should almost instantly pressurize. The simulation evidently can't determine a seal or assumes a separate grid cannot be part of a seal. This is too bad because it limits the size of a hangar door.
Regards,
Bill
Replies have been locked on this page!