Experimentation of insulation values of walls and doors.

[top_hat_tomato]

So... A few months ago I read a short post describing the efficient ways to make walls in Rimworld with information like 'corners don't matter' and '2 thick walls are the best' so I decided to make this post to re-evaluate the information.

Preface - All measurements were taken with an external temperature of -270 c , wall material is irrelevant, and each room started at 2000 c at the same time.


(Graph is Temp (in c) by Time (in-game ticks)
Conclusion : Corners do not in fact matter, doors impact insulation greatly, and double walls do indeed help quite a bit.

Tests were done with a constructed roof. Having a mountain overhead seems to be more effective, but needs to be tested more.



Data : https://docs.google.com/spreadsheets/d/1JZSr7q4TXLGtAvVN9QrpZ21m5bPN3Qf9QS80G2LIp0I/edit?usp=sharing

[O Negative]

If you don't mind me asking,

What method did you use for data collection?
What about equation derivation?
I see numbers and graphs that I want to believe, but I can't readily do so unless I understand exactly how they were obtained. Silly-skeptical-ol' me Lol

[top_hat_tomato]

If you don't mind me asking,

What method did you use for data collection?
What about equation derivation?
I see numbers and graphs that I want to believe, but I can't readily do so unless I understand exactly how they were obtained. Silly-skeptical-ol' me Lol

For data collection I used many debug 'god' powers. Use cold snap until temperature is -270 (as that is the minimum in rimworld). Next I used the 'build roof' tool. Then the 'push heat 1000' which made every room immediatly 2000c (the upper temperature limit). Then I used the tick counter in another debug menu to capture time. The reason that the tick numbers are not equidistant is because I am unable to use programs such as cheat engine to slow down the game.

For your second statement:
I used exponential decay as the parent function (y) = (initial) * (rate)^x . I used that function because it allows me to have a horizontal asymptotic and it seemed the most logical step because Δtemp would have less of a change over time (for more definite reasoning, go to https://en.wikipedia.org/wiki/Exponential_decay and ctrl + f (search) for Heat transfer on that page, it states

"  If an object at one temperature is exposed to a medium of another temperature, the temperature difference between the object and the medium follows exponential decay (in the limit of slow processes; equivalent to "good" heat conduction inside the object, so that its temperature remains relatively uniform through its volume). "

(also, the data follows a curve that is backing this inference up)

If you do find any errors, please do tell. I actually want to have this be as accurate as possible.

[brcruchairman]

Bravo, sir. Bravo. *salutes* I really appreciate you gathering and defining precise data (e.g., temperature and tick pairs, temperature curves, all for various configurations.) The only thing I might like would be pictures of the setup, though upon careful examination of (read: actually clicking on the link to) your data, the verbal descriptions were clear enough for replication.

*applauds* Well done, sir. You rock.

[top_hat_tomato]

Bravo, sir. Bravo. *salutes* I really appreciate you gathering and defining precise data (e.g., temperature and tick pairs, temperature curves, all for various configurations.) The only thing I might like would be pictures of the setup, though upon careful examination of (read: actually clicking on the link to) your data, the verbal descriptions were clear enough for replication.

*applauds* Well done, sir. You rock.

What do you mean by pictures of the setup? I believe that they're part of the image provided (the embedded one with the graph and formulas) or is that not loading for you? ( http://i.imgur.com/gT0tPCx.png )

[brcruchairman]

<.< Adeeeerp. My bad. I was just looking at the graph and text, and completely skipped over the embedded pics. I withdraw my only complaint; you, sir, are awesome! *offers high-five*

[O Negative]

For data collection I used many debug 'god' powers. Use cold snap until temperature is -270 (as that is the minimum in rimworld). Next I used the 'build roof' tool. Then the 'push heat 1000' which made every room immediatly 2000c (the upper temperature limit). Then I used the tick counter in another debug menu to capture time. The reason that the tick numbers are not equidistant is because I am unable to use programs such as cheat engine to slow down the game.

For your second statement:
I used exponential decay as the parent function (y) = (initial) * (rate)^x . I used that function because it allows me to have a horizontal asymptotic and it seemed the most logical step because Δtemp would have less of a change over time (for more definite reasoning, go to https://en.wikipedia.org/wiki/Exponential_decay and ctrl + f (search) for Heat transfer on that page, it states

"  If an object at one temperature is exposed to a medium of another temperature, the temperature difference between the object and the medium follows exponential decay (in the limit of slow processes; equivalent to "good" heat conduction inside the object, so that its temperature remains relatively uniform through its volume). "

(also, the data follows a curve that is backing this inference up)

If you do find any errors, please do tell. I actually want to have this be as accurate as possible.

Hm, I'd like to do a second study as well, actually. Just to confirm your results, if you don't mind.

I like your use of an exponential decay model, which may accurately reflect how temperature is coded. It even does a decent job of simulating actual Thermodynamics laws. However, I'd like to go through some of my old Physics - Thermodynamics notes to see if it holds up in a logical way

For now, good job on this, friend! Only recommendation would be to rearrange the legend/key on the left hand side to follow an ascending or descending insulation rating (if that makes sense)

[ArguedPiano]

Thank you Top_hat_tomato, this is quite helpful.

[Spingitore]

Great work, this is both 'scientifically' interesting and helpful for game decisions.
Yet, being a newbie to this game, I am not really sure if it is worth building thick walls. Even looking at these numbers, isn't just easier to build one additional power source to maintain temperatures inside your colony instead of designing a quite complex structure with double external walls?
I guess this second option is much cheaper in terms of material, but what about the time, both in real time and in colonists work, to plan and eventually rebuild everything?
Maybe a good compromise is to build thick walls only in storage areas outside the main building (assuming the main building is the residential one, that's what I do actually).
And maybe, just maybe, add double walls and double doors only to the entrance of main building (which sometimes can be a bit tricky anyway, depending how you designed the building). I am not sure, however, if that would be a great benefit and maybe would need additional testing?

[top_hat_tomato]

Great work, this is both 'scientifically' interesting and helpful for game decisions.
Yet, being a newbie to this game, I am not really sure if it is worth building thick walls. Even looking at these numbers, isn't just easier to build one additional power source to maintain temperatures inside your colony instead of designing a quite complex structure with double external walls?
I guess this second option is much cheaper in terms of material, but what about the time, both in real time and in colonists work, to plan and eventually rebuild everything?
Maybe a good compromise is to build thick walls only in storage areas outside the main building (assuming the main building is the residential one, that's what I do actually).
And maybe, just maybe, add double walls and double doors only to the entrance of main building (which sometimes can be a bit tricky anyway, depending how you designed the building). I am not sure, however, if that would be a great benefit and maybe would need additional testing?

This information is only actually useful for extremely cold areas ( ex. -70c average ) or if you have a mod like Marv's disasters (which adds constant world conditions like rogue planet and solar apocalypse where the temperature gets colder and colder/warmer and warmer).

I'm also considering using a system of equations to figure out a function for different room sizes and configurations.

[Spingitore]

Great work, this is both 'scientifically' interesting and helpful for game decisions.
Yet, being a newbie to this game, I am not really sure if it is worth building thick walls. Even looking at these numbers, isn't just easier to build one additional power source to maintain temperatures inside your colony instead of designing a quite complex structure with double external walls?
I guess this second option is much cheaper in terms of material, but what about the time, both in real time and in colonists work, to plan and eventually rebuild everything?
Maybe a good compromise is to build thick walls only in storage areas outside the main building (assuming the main building is the residential one, that's what I do actually).
And maybe, just maybe, add double walls and double doors only to the entrance of main building (which sometimes can be a bit tricky anyway, depending how you designed the building). I am not sure, however, if that would be a great benefit and maybe would need additional testing?

This information is only actually useful for extremely cold areas ( ex. -70c average ) or if you have a mod like Marv's disasters (which adds constant world conditions like rogue planet and solar apocalypse where the temperature gets colder and colder/warmer and warmer).

I'm also considering using a system of equations to figure out a function for different room sizes and configurations.

So double walls are never worth it in standard conditions? Even partial double walls/doors?

That would be interesting, my advice is to use different colors from the ones you used here as I struggled a bit to mentally link the curves to the legend.

Again, good job!

[Zhentar]

I wouldn't say they're never worth it, but I do think people overuse them. IMO, they are most useful with blocks of small bedrooms (where they can potentially save you from needing to put a vent or heater in every single bedroom), or when you need to tolerate power interruptions (freezer in hot maps, farms in cold maps).


top_hat_tomato: if you're going to try to make a general equation, you should know that due to some rounding behavior, small rooms can lose heat out the walls a fair bit faster than they should (close to 50% more in the worst case).

[top_hat_tomato]

top_hat_tomato: if you're going to try to make a general equation, you should know that due to some rounding behavior, small rooms can lose heat out the walls a fair bit faster than they should (close to 50% more in the worst case).

I actually think that behavior may have to do with the walls to room area ratio. Now, I haven't done any testing on it so this is just a theory.

A 1 by 1 room (1u area) is surrounded by 4 walls ( 1:4 ratio) while a 4 by 4 room (16u area) is surrounded by 16walls (1:1 ratio).

But also to your point, I would also have to use some approximations on my end to get a system of equations to work with multiple trials.