Dwarf Fortress Bug Tracker - Dwarf Fortress
View Issue Details
0002676Dwarf FortressCreaturespublic2010-07-13 09:142012-02-21 13:41
0002676: Fatty tissue has same melting point as rendered fat/tallow, which is far too low
Fatty tissue in the game uses the "fat" template, which has very similar properties to rendered fat (the "tallow" template). This does not lead to the desired (or realistic) behavior, mostly because fatty tissue in real life does not consist solely of its constituent fats; it's part of a cellular/tissue matrix.

The implicit assumption in the raws right now is that fatty tissue should melt when its constituent fatty acids melt. In the raws, this is 110 degrees Fahrenheit. As a simple counter-argument, many/most cells in the body are mostly water, but don't melt away and slough off at water's melting point. In addition, in the human body (and presumably many animal bodies), much fatty tissue is composed of fat that melts well below human body temperature. In fact, according to some research I've done, the fats in the human foot are largely liquid at below the melting point of water. As another possible counter-argument, if I go heat up a steak right now, there will still be fat/marbling on it when I'm done, even where it is exposed, despite the fact that a temperature far in excess of what is necessary to melt the fats themselves has been achieved.

At any rate, this appears to be one of the primary causes of the Everybody Melts type bugs. Obviously, there are other issues involved (like ambient/environmental temperature and how well bodies insulate/regulate their temperature), but fatty tissue should never liquefy at such a low temperature.

I've had trouble determining more realistic values because adipose tissue varies so wildly, doesn't necessarily undergo a very clear-cut phase change, and due to a lack of available data in the first place, not to mention the fact that the situation changes once the material is pressed between other tissues. From what little I've been able to determine, I think a value of at least 140 degrees Fahrenheit is necessary, but possibly higher. This might not seem like a huge difference, but when it comes to environmental circumstances and body temperature, it is. After all, we don't want dwarves melting just because they have a high fever and live in Costa Rica.
Is always the case. Effects may be observed by submerging an animal in magma or otherwise subjecting them to a source of high heat: Often, the fatty tissue has melted out of the creature before (m)any other wounds have developed.
There are some other raw values for body tissue materials that deserve changing (such as lowering the HEATDAM_POINTs to simulate tissue damage that occurs at high temperatures), but this is the only one I consider so far to lead to what I'd call "buggy" behavior.
fat, material, melt, melting, raw file, raw files, temperature, tissue layers, tissues
related to 0002605new  Creatures bleeding to death after water coverings evaporate from small body parts (toes, lips) 
related to 0000263new  No announcement when fat melts away 
Issue History
2010-07-13 09:14G-FlexNew Issue
2010-07-13 09:48smjjamesNote Added: 0010149
2010-07-13 09:50smjjamesNote Edited: 0010149bug_revision_view_page.php?bugnote_id=0010149#r3937
2010-07-13 12:52FootkerchiefRelationship addedrelated to 0002605
2010-07-13 12:53FootkerchiefRelationship addedrelated to 0000263
2010-07-14 01:18G-FlexNote Added: 0010222
2010-07-14 01:20G-FlexTag Attached: fat
2010-07-14 01:20G-FlexTag Attached: melt
2010-07-14 01:20G-FlexTag Attached: melting
2010-07-14 01:20G-FlexTag Attached: temperature
2010-07-14 01:20G-FlexTag Attached: tissues
2010-07-14 01:20G-FlexTag Attached: tissue layers
2010-07-14 01:27G-FlexNote Added: 0010224
2010-11-30 04:05SanDiegoNote Added: 0014309
2010-11-30 10:10FootkerchiefNote Added: 0014312
2010-11-30 17:00G-FlexNote Added: 0014323
2010-11-30 18:47G-FlexNote Edited: 0014323bug_revision_view_page.php?bugnote_id=0014323#r5455
2010-12-01 11:25SanDiegoNote Added: 0014349
2010-12-01 19:44Hieronymous AlloyIssue Monitored: Hieronymous Alloy
2010-12-01 21:25tatterdemalianNote Added: 0014357
2010-12-01 21:36G-FlexNote Added: 0014358
2010-12-01 21:42G-FlexNote Edited: 0014358bug_revision_view_page.php?bugnote_id=0014358#r5467
2011-01-10 20:38dwarf_sadistNote Added: 0014841
2011-01-19 00:38G-FlexNote Added: 0014908
2012-02-21 13:41G-FlexTag Attached: material
2012-02-21 13:41G-FlexTag Attached: raw file
2012-02-21 13:41G-FlexTag Attached: raw files
2016-03-03 13:47ButtonIssue Monitored: Button

2010-07-13 09:48   
(edited on: 2010-07-13 09:50)
Hey G-Flex, this one would be the cause behind the bug that I've observed 0002605 . This would also serve as a good example for your case.

To note, the temperature that I observed in that savegame is 10080 (or 10081), which converted into fahrenheight (thanks to the wiki), is 112 degrees. While that temperature would certainly give people problems if they aren't prepared, you certainly don't see peoples fat melting in the Sahara or Death Valley.

2010-07-14 01:18   
Yeah. To be fair, the body should regulate temperature well enough that the layers underneath skin shouldn't reach 110 degrees to begin with; it should be kept closer to core body temp in the first place.

That (and the fact that being covered in water somehow causes you to become hotter instead of cooler!) seems to be a distinct bug, but yeah, this is all definitely interrelated.
2010-07-14 01:27   
As a side-note, I have to wonder how hot a dwarf's *brain* gets in situations like that. I have a suspicion that Dwarf Fortress doesn't really care much if a dude's core body temperature goes way out of range.

I say this because I've gotten animals with _burn wounds on their brains_ in Arena Mode at least a couple times before. Before they die. The heatdam_point for brain material is 282 degrees Fahrenheit, so, er, slight problem there. It should at least simulate whatever fever effects would occur at that temperature (or better yet, fevers should simply raise the regulated body temperature, since that's what they are, and the effects should be linked to temperature itself).
2010-11-30 04:05   
I was alerted to this issue in my Tissue Rebalance thread and if I may I'll say a few things:
1) Fatty tissue consists mostly from fat. The fatty cells (adipocytes) are really just thin shells enveloping blobs of fat. Then there are some collagen fibers and some blood vessels. But it's mostly fat.
2) Tallow is term used for fat coming from cattle (be it cows or sheep). There is slight chemical difference from other fats, lard for example, in ratio of saturated fatty acids to non saturated ones. This causes it to be slightly stiffer than other fats. But essentially they are the same.
3) Therefore, I see nothing wrong with fat and tallow materials being similar.
2010-11-30 10:10   
I'd like to see a citation for the claim that blood vessels in adipose tissue don't allow significant heat exchange.
2010-11-30 17:00   
(edited on: 2010-11-30 18:47)
SanDiego: I disagree, for a few reasons.

1) "Tallow" in DF does not refer to fat from a particular form of animal. I know it usually does in real life, but that's not what we're talking about here. In DF, "tallow" is pure, rendered fat, whereas "fat" is the raw material existing inside the animal's body and extracted from it. The reason "tallow" is used is because there's no general term for rendered fat, only terms like "tallow" which refer to rendered fats from certain animals. I would think you would know this if you're making a body material mod.
2) I know fatty tissue is mostly fat. It doesn't matter what it "mostly" is; what matters is the properties of the material. A balloon is mostly air surrounded by an incredibly thin shell of solid material, but that doesn't mean a cluster of balloons evaporates at room temperature. A very large cluster of tiny balloons is still "solid". Those "thin shells" matter here. All you have to do is heat up a steak to realize that fatty tissue doesn't melt away the second the fats themselves reach their melting point. It gets softer, and oozes some stuff out, but it doesn't fall apart or melt away. This is trivially demonstrable and something we experience on a regular basis in everyday life. You have to realize that the material properties in DF aren't the properties of the majority chemical constituent, but the properties of the material as a whole.

Using the logic of the game, and that you're using (that fatty tissue should melt at the same temperature as its fatty acid constituents), some fats in the human body should be melting out even during freezing weather. Obviously, this isn't and shouldn't be the case, so something is wrong.

2010-12-01 11:25   
G-Flex: I was talking purely about materials. Rendered or not, fat is still fat - very long chains of carbon with some -OH group at the other side. The reason fatty tissue does not melt (under normal circumstances like not being on fire) is that human body has really astonishing thermoregulation.
If I remember my high school chemistry lessons well, rendering fat only increases the ratio of saturated acids, but it hardly has any effect on calorimetric atributes - only how solid it is (by the way - margarine is made the very same way from plant oil).
In short - tallow vs. fat, tallow is more "solid" and fat more "liquid" but for any other purposes they are pretty much same. I have excessively thick Biochemistry textbooks if you'd like to know more on the subject.
2010-12-01 21:25   
The rendering process separates the fatty acids from the proteins, water, and other compounds present in the tissue. It has nothing to do with increasing the ratio of saturated and unsaturated fats, that's an entirely different process called hydrogenation.

Fats are typically liquid at body temperature, and varying degrees of solid at room temperature. Fatty tissue, like G-Flex says, is actually fats surrounded with protein shells, that rapidly break down when removed from a living organism, but the remnants of protein and the water within the cells remain, not only making the fatty tissue more tough and unpredictable in chemical reactions, but also inducing rapid spoilage (microorganisms thrive better on food + water than on food alone, which is also why candifying helps preserve food even though one would think the sugar would encourage bacterial/yeast growth, not retard it).

The main point is, G-Flex is quite right. The melting point of fatty tissue, especially within the body, should be the melting/burning point of the proteins surrounding the fat, not the melting point of the fat itself. The rendering process takes place at a lower temperature than would be necessary to melt the protein, because the point is to separate the protein from the fat, not melt it all into one big mass of hydrocarbons.
2010-12-01 21:36   
(edited on: 2010-12-01 21:42)
"-Flex: I was talking purely about materials. Rendered or not, fat is still fat - very long chains of carbon with some -OH group at the other side. The reason fatty tissue does not melt (under normal circumstances like not being on fire) is that human body has really astonishing thermoregulation."

You're still mistaken here.

1) The "material" in question is not pure fat and never was. It's fatty tissue, period. You cannot take the primary constituent of that material and pretend that it's all that constituent, or else you have to pretend that brains are made out of water and melt out your ears at the slightest provocation.

2) It has nothing to do with thermoregulation. The fat in your feet melts below zero degrees celsius, and I highly doubt your feet are at zero degrees celsius, or any part of your feet for that matter.

"Fat is still fat" is a useless thing to say here, because it's not just fat, and because even if it were "just fat", different fats have very different melting points, even within the fatty tissue of the same animal.

2011-01-10 20:38   
In order to fix the issue of fat not melting away the instant it reaches the melting point is to do with physics. In order to make a REALISTIC temperature effect, you would have to overhaul the heat hard codes similar to how combat was overhauled.

Right before boiling begins, the object must gain enough heat in order to overcome it's latent heat. Once a substance reached the boiling point, it stopped gaining heat and starts to boil. The more latent heat is added, the more of the substance can turn into a gas(or a liquid if going from a solid) After this point, adding more heat would raise the temperature of the product, such as steam produced by heating water.

Implementing this into DF would allow you to customize how fast a substance changes states. Fat could have a very high latent heat threshold, meaning that it wouldn't all melt away instantly, unless the temperature was very high.
2011-01-19 00:38   
dwarf sadist: Heat of phase change is important, yeah, but it's honestly not the main issue here. The more pertinent issue is the melting point itself.