Print

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

[Zucchini]

Hi all.

I'm working on a set of real-world physics numbers for my own use and for anyone else who would see fit to use them, possibly even for submission to Toady.

But I'm not in the science department.  I'm in the history department.  Yes, Martha, I'm not the history department, I'm in the history department.  And I need someone in the math biology department.

(10 points to anyone who gets that reference...  )

But actually, sort of the crossroads of material science/physics and biology departments.

So, being something of a SIMP, I'm going to need help.  Can you people literate in the appropriate areas give me a bit of advice?

First problem:

I want to come up with real-world Yield, Fracture and Strain at Yield values for bone, skin, and other keratinous, collagenous and bone-based biomaterials.  Unfortunately, the viscoelastics are nowhere near as straightforward to read as inorganics.  And if I use the "yield" value most papers give, it'll distort the results such that organics are brittle.

So, the question is, how do you read an appropriate, relevant Yield value from the stress/strain diagram of a viscoelastic material?

Note that, since in DF we're talking about axe surgery and not plastic surgery, we're really not concerned about the lower yield-to-viscoelastic deformation "yield" that most of the literature talks about, but rather to the yield-to-beginning-of-failure that is more like inorganic yield (i.e., where the material is strained past the point of its ability to elastically deform, therefore straining plastically and damaging it).

But the stress/strain diagrams and the literature really don't give clear guidance on this.  At all.  So, as a matter of the question applied to concrete example, what Yield values should be estimated from visually reading the stress/strain diagram below?

Eyeballing it, here's what I got, but I'm really not very confident at all with my understanding of what we should treat as the Yield point for collagenous/viscoelastic stuff:

Tendon: 16.5 MPa or so (from where I see the 'crook' in the curve before it goes linear up until failure)
Rhinoceros Skin: about 25 MPa?  (from about where the direction of the curve changes from concave to convex)
Cat Skin: about 8 MPa

Am I totally off on these?






This is a sort of repost of the question from the modding forum.  And as I said there, I will be obscenely grateful for your help.  And as I clarified there, I mean grateful to an obscene degree, not in an obscene way.  (!)

[Detahramet]

But actually, sort of the crossroads of material science/physics and biology departments.

...since when did biology stop being a science?

[SquatchHammer]

But actually, sort of the crossroads of material science/physics and biology departments.

...since when did biology stop being a science?

Read what you quote. Material Science specifies certain materials whether being organic or inorganic.

Since he's into the history he doesn't have the specific knowledge needed to put in the paticular numbers for any organic material ie bone, skins, other tissue.

[Detahramet]

Ah. So I'm an obscenely massive idiot then. Sorry for my incompetence.

[SquatchHammer]

Nah its the old saying Look before you leap. Or the Dwarfish saying have a very good exit when you open up the magma for the forges.

[oven_baked]

But actually, sort of the crossroads of material science/physics and biology departments.

...since when did biology stop being a science?

Ask any chemist or physicist and they will agree that biology is not a science.
(But either way he's talking about the crossroad between Material_science/Physics and Biology)

[Lich180]

Ask any chemist or physicist and they will agree that biology is not a science.
(But either way he's talking about the crossroad between Material_science/Physics and Biology)

Spoiler (click to show/hide)

Found it relevant. Thank you, xkcd.

[Shinziril]

I suppose I'm probably qualified to answer this, given that I actually have a degree in Materials Science now . . .

The 0.2% offset method of calculating yield strength is commonly used for metals.  Going by that test (but mostly my own eyeballs), the graph for tendon never really yields at all, and the graph for rhinoceros skin either doesn't yield or yields just barely before failing.  The graph for cat skin looks more like the kind of stress-strain curve you'd get off an elastomer (rubber), which can't really use the same framework of elastic modulus/yield strength/ultimate strength that we use for metals. 

Note: in stress-strain plots from actual experiments, there's often a little non-linearity at very low stresses (e.g. the plot for rhinoceros skin).  You can usually just ignore that- sometimes it's from slack being taken up in the testing apparatus, or maybe there's other reasons, but it will usually switch almost immediately to proper linear elasticity. 

I'm trying to come up with a good way to phrase this, but I basically suspect that for a lot of skin types there's really no sensible way to fit them into the standard elastic modulus/yield strength/ultimate strength model used in DF (and a lot of actual engineering).  That model works well as a very simple model for metals and ceramics (ceramics and other brittle materials obviously having a yield strength equal to their ultimate strength), but squishy polymers (and fiber-strengthened soft tissue) just don't react to stress the same way.  If you define "yield strength" as the stress at the onset of plastic deformation, it gets even worse, since it's quite possible that an elastomer (or, as the case may be, cat skin) doesn't permanently deform at all until it fails completely (at least in a short-term test, but if you want to be more accurate we have to get into slow-deformation-over-time processes like creep and stress relaxation that aren't even used in DF at the moment and are even more annoying and aargh I'm probably putting way too much thought into this). 

[squishynoob]

The problem is that when coding the combat engine, I believe Toady has cut corners regarding elasticity. It becomes evident if you fiddle with the raws enough.

For instance, you know that blunt weapons, no matter how much force they have behind, never manage to tear, or completely crush, a soft elastic tissue. Even when stomped on by a giant, you'll only receive bruises and in the worst case broken bones.
Now, in the material definition raws, you'll run across elasticity values, in case of blunt damage, IMPACT_STRAIN_AT_YIELD.
For each of the soft tissues the value is 50000.
Now change them to 49999. That's a 0.002% change. It won't do anything meaningful, right? Fire up the arena, and get ready for a big surprise: each and every punch or kick will now tear soft tissues! At this point it becomes obvious that bruising or tearing is merely determined by a on/off threshold, and not a curve of sorts.
Also, if the attack doesn't have enough force to tear the tissues (I believe, it is between the "yield" and "fracture" values), they are "dented", which appears somewhat different from bruising - it causes pain, for one.

Another example is chain mail. Unless a harder metal is used for the weapon, pretty much every edged attack will be converted to blunt. Yes, even the overpowered railgun slugs bolts that punch right through giant-sized plate.

[Zucchini]

Oh my. 

Apologies for the apparent snub.  It's just that some fields have "science" in their name, and some don't, and it has nothing to do with whether they're actually scientific or not.

Witness "political science."  "Lol."

"Material science" is the actual name of the field as far as I've been made aware -- it wasn't meant to imply physics or biology are not scientific.  They just don't have it in their name. 


Tangent: Funny but true anecdote on the difference between "political science" and history students

Spoiler (click to show/hide)

I majored in history and almost minored in poli sci, so I took classes in both.  I noticed something interesting in how the poli sci majors and the history majors differ in their approach to the reading.

In my history classes, if someone had not done the reading for the week, they would hide in the back of the room (or carefully in the middle if they suspected that the back would be too obvious) and would be very, very careful not to call attention to themselves.  They would very, very sheepishly confess that they had not done the reading if called on, and appeared to be highly embarrassed.

So, we could say that the amount of reading done was inversely proportional to the student's willingness to pipe in on class discussions about the material.

Now switch to poli sci classes.

Some dude hasn't done the reading?  That's the guy that's going to be dominating the in-class discussion for the day.

I noticed that at least half the class hasn't done the reading on any day, and the ones who had done the reading were actually less likely to speak up.  No.  Actually reading about the issues would complicate things.  It was the ones who had not done the reading who were much more likely to speak up, and were likely to just spout out whatever they already thought about the matter -- usually some kind of opinion that brings to mind the old truism about the equivalent ubiquity of opinions and anuses.

The teacher was too tired and beat down to contain it, and would patiently let them finish and either hope someone who had done the reading would speak up, or just move on.

So the likelihood of boldly speaking in class for poli sci students was inversely proportional to the amount of the assigned reading they had done.  And they were bold, brash and invincibly confident of what they had to say.

I've had this opinion verified by others who've had the same experience at different colleges.

Yay political science.

Well, I do recommend that you look at:

[link to SanDiego's Tissue Rebalance mod]

due to its effectiveness at re-working bone, tissue, and other biological matter structures to much more appropriate levels.

Oh yeah, I use it in my own build.  Great stuff!

I suppose I'm probably qualified to answer this, given that I actually have a degree in Materials Science now . . .

The 0.2% offset method of calculating yield strength is commonly used for metals.  Going by that test (but mostly my own eyeballs), the graph for tendon never really yields at all, and the graph for rhinoceros skin either doesn't yield or yields just barely before failing.  The graph for cat skin looks more like the kind of stress-strain curve you'd get off an elastomer (rubber), which can't really use the same framework of elastic modulus/yield strength/ultimate strength that we use for metals. 

Note: in stress-strain plots from actual experiments, there's often a little non-linearity at very low stresses (e.g. the plot for rhinoceros skin).  You can usually just ignore that- sometimes it's from slack being taken up in the testing apparatus, or maybe there's other reasons, but it will usually switch almost immediately to proper linear elasticity. 

I'm trying to come up with a good way to phrase this, but I basically suspect that for a lot of skin types there's really no sensible way to fit them into the standard elastic modulus/yield strength/ultimate strength model used in DF (and a lot of actual engineering).  That model works well as a very simple model for metals and ceramics (ceramics and other brittle materials obviously having a yield strength equal to their ultimate strength), but squishy polymers (and fiber-strengthened soft tissue) just don't react to stress the same way.  If you define "yield strength" as the stress at the onset of plastic deformation, it gets even worse, since it's quite possible that an elastomer (or, as the case may be, cat skin) doesn't permanently deform at all until it fails completely (at least in a short-term test, but if you want to be more accurate we have to get into slow-deformation-over-time processes like creep and stress relaxation that aren't even used in DF at the moment and are even more annoying and aargh I'm probably putting way too much thought into this). 

Nonono, huge thanks!  That's exactly the kind of advice I need on it, so much appreciated.  I feel I have some better guidance for eyeballing DF-workable "Yield" numbers now (despite the problematic nature of the whole thing that you and Squishynoob describe).

The problem is that when coding the combat engine, I believe Toady has cut corners regarding elasticity. It becomes evident if you fiddle with the raws enough. [ . . . ]

Yep, I've been made aware of the how DF's framework really doesn't handle organics well (because, as you guys point out, it treats them all like inorganics, and the whole thing on elasticity/yield).  However, that specific advice you give will be very helpful as well, so thanks.  With all the relevant info shotgunned all over the forum, it's nice to have it crystallized like that.

Here's to hoping Toady implements the physics better!  I suppose it may be a waste of time, but if he does ever improve them, this research may come in useful.  Until then, though, the short-term less-ambitious intent for it is to be the basis for skin/leather variation.  I'll have to figure out exactly how much I can reasonably vary this or that in terms of DF's current paradigm once I've got a reasonable set of numbers to work with.

@Lich180 -- I've never seen that graphic.  It's great!  Just need to get poli sci on there...