Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Zucchini]

Oh, and a question for the engineers/more-physics-literate-people here:

What is strain at yield a direct expression of?  EDIT AGAIN: Answered. I'll bring it up in a new thread for discussion, because that didn't have to be so hard.  At all.

But in case curious, the answer is:

Spoiler (click to show/hide)

Strain at yield is literally the same value as seen on stress/strain diagrams: it's the percentage of material deformation (i.e., the stretching or compressing or twisting or whatever) that occurs at the point where the material has been stressed to the limit of its ability to deform elastically (i.e., in such a way that it can return to its original form with no structural damage), and any further stress (force applied to it) will cause it to deform plastically (in such a way that its structure changes permanently -- i.e., is damaged).  On the stress/strain curve diagrams, strain at yield is expressed as a percentage of the material sample's dimensions.  DF simply multiplies this by 100,000 for its SAY value.

So, 1% deformation (strain) at yield = 1000.  DF's skin SAY value of 50000 equates to 50% deformation -- at the yield point, according to vanilla values, 1 cm of skin can stretch to 1.5 cm long without damage.  If you let it go, it'll go back to 1 cm long.  If you keep applying stress, it's gonna get damaged. Incidentally, this value is way too high -- see example 2 for why.

So, to determine an appropriate SAY value for DF's purposes, all you need is either (a) a stress/strain curve diagram, or (b) the yield stress value in MPa and the Young's modulus in GPa -- divide yield by YM, and you have your SAY.  If (b) is the case, then just divide the yield stress by YM x 100,000; the result is your percentage strain at the point of yield in real-world terms. You then multiple that value by 10,000, and you have a DF strain at yield value.

Example 1: In the vanilla raws, Iron's tensile yield is 155000 MPa. Its Young's Modulus is 211 GPa.  1 GPa (gigapascal) is 1,000 MPa (megapascal).  So, you divide 155000 by (211 x 1,000,000) and you get 0.00073, which is your real-world percentage (0.073%) deformation of iron in the tensile stress test machinery.  Multiply this by 100,000, and you get 73 -- the value Toady used for SAY for iron.

Example 2: Skin. Looking it up, we can find skin has a Yield of around 15 MPa and an elastic modulus (same thing as young's modulus except when specified otherwise) of 83.3 MPa (note the M -- here, YM is in MPa, not in GPA, so we'll need to multiply YM by 1,000 instead of by 1,000,000).  So, 15000 / 83,300 = about 18.01% deformation at the yielding point.  Multiply this by 100,000 and we get a 18,000 for our DF SAY value -- or 18,010 if you want to be anal about it.


This process can be shortcutted to Yield / (YM * 10) = Strain at Yield in DF terms, if you'd prefer not to first obtain the real-world SAY value. 

Further note: All DF stress values are measured in KPa. What we're really wanting to do is convert YM to KPa (kilopascals). So, as in example 2 where we have a substance like skin with a YM measured in MPa instead of GPa, we'll only multiple YM by 1,000 before dividing yield stress by it.

Original post asking it really stupidly:

Spoiler (click to show/hide)

I'm just barely starting to get a handle on what all this physicsy stuff means, and I'm wondering if strain at yield is a simple expression of strain on the stress-strain plots (say, multiplied by something such that 100000 = 100% strain).  That's the impression I'm getting, since the highest S.A.Y. in the game seems to be 100000, though I imagine it could also mean 1000%.

First edit, asking it in a slightly more intelligent way after a bit more research:

Spoiler (click to show/hide)

According to this (and consistent with what I've seen in other sources), most metals tend to score around 0.01% SAY on tensile tests.  If my impression that SAY as used in the DF raws is a simple, literal expression of strain at the yielding point, then it seems reasonable to assume that Toady's SAY value for metals = 100 = 0.01% = 0.0001.  Thus, antler, which starts to suffer damage around 0.05% (according to one source; still researching this, but let's assume so for argument), would accordingly be around 500.  Skin's score of 50000, according to this standard, would be 5% (50000) elastic deformation before structural damage begins, and scale's score would be 10% (100000).

Do I have the right sense of this, or am I totally off?  It seems too simple, like somehow I'm supposed to have to run it through some complicated equation (or at least way over my head) involving Young's modulus, but a vegetable can hope.

[Grimlocke]

General body hair on furry creatures becomes kind of confusing. What I have done before, was to make sure the tissue was still called 'skin', and left the hair tissues where they are to prevent the headache of non-fur hair being weird, and having to remove all the tissues by hand. So far I have not found a way to prevent hair from showing up at all after butchery, or a way to keep cetain tissue layers together after butchery.

Yeah, I'm not quite sure what to do here.  My inclination thus far has been to go all the way and change the methodology to integrate fur and skin in the same layer, just getting rid of hair as a tissue layer except where it makes specific sense (manes, furry tails, etc.).

I'm not at all sure that it wouldn't be better to keep "fur" as a separate tissue layer, like it's already set up with hair by default, but that nonsensical butchering product is bothersome.  Doesn't make sense to have a butchered ermine yield both "ermine pelt" and "ermine fur"/"ermine hair."  On the other hand, something like a horse would technically look more like a human -- a creature with a normal skin that drops hair (mane and tail).

Looking at your experience, if I understand it all correctly, to do it this way will get away from that problem of "hair" showing up in butchering products, but it's going to involve a hell of a lot of manual removal of hair from creatures.  But it's the only non-awkward way to do it -- and nonsensical butchering products fit into that "workaroundiness" I'm trying so hard to get away from.  Hrm.  Lots of work.

I also notice you added a seperate material for the tissue layer fur. What you can do, is to simply use the tissue layer fur for the tanned items as well so that you only need to set one color per creature.

Wouldn't that mean that the pelt/furred skin wouldn't need to be tanned?

It would still need the tanning reaction in order to be turned from bodypart into tanned skin item. This also stops rotting, from gameplay perspective its essentialy the same result.

As for FBs, from what I know they are built basicaly the same as vanilla creatures. Standard mats+tissues, + whatever body detail plan their specific nature requires, like the hair body detail plan, humanoid relsizes, ribcages etc etc. The standard mats will allways load first, so if you modify another body detail plan to overwrite one of the materials it should remove the original. It is a somewhat sloppy way to do it, but right now also the only way and the end result should be good.

The question would be, then, which BDPs to modify...  But wait.  I think I'm seeing it...  So, for example, for hardcoded vermin birds, take a later BDP that the game will always apply to them -- say, one of the new feathered body part ones -- and use it to remove the "monstrous"-strength tissue layers and replace them hollow bones and weak skin and gizzards, etc.?  (Or remove all specific organs in favor of nonspecific bird viscera.)  If I've got what you mean right, AWESOME -- that would solve the bird/FB hard-coding BDP problem.

Body detail plans can add materials, tissues, tissue layers, and bodypart relations (like skulls being around brains). Sadly you cant remove bodyparts, and while you can make an additional copy of the vertebrate tissue layer plan to use just one material for all the organs, that would still mean you have to go past each vermin bird and change that part.

Changing the feather tissue layer plan would probably also cause trouble with feathery FBs. Its probabaly best left as it is. Removing tissue layer hair however should be fairly easy. There are seperate plans for body-covering hair and just head hair like humans have. Just replace the first one with [ADD_TISSUE:SKIN:FUR_TEMPLATE], and you have a creature without hair or skin tissue layer and just fur instead.

Note that all lines used by the random creatures generator can be found in the string dump. Maybe someone has made a specific article about it by now, im not sure. String dump link: http://dwarffortresswiki.org/index.php/String_dump

...  which also suggests I may want to relegate the CHITIN_MATERIALS and CHITIN_TISSUES to FB-only use like I did for the STANDARD ones...

I think FBs use allmost every body detail plan, so all the material plans will need to be kept for them only.

[Zucchini]

Alright, thanks Grim.  I'll be a lot more intelligent about this stuff once I wade into it, and then I'm sure I'll be a lot better able to digest all of it.

---

Spent the last four days trying to figure out just what the hell all these physical data like strain at yield really mean, so I'm now I'm going back through all my numbers and coming up with final numbers for bone, horn, tooth, etc.  Should be done with that in a few days, and then I'll be getting into the BDPs in earnest.