478 replies to this topic

[Y E O N N E]

Nightbird, on 05 March 2018 - 04:20 PM, said:

I don't believe PGI ever said that they scaled by volume of any sort, this PSA is more or less targeted at people on the forums that are under the impression PGI did. This is not the first thread to try to debunk this impression

They did.

They took a cubic measurement of every 'Mech, and matched 'Mechs of the same weight to a single cubic value, then made tweaks from there by feel. What they didn't do was take a single value to anchor the scaling across all weights, which is more or less what I said to you last time.

[Nightbird]

So I got a question. I've been thinking though the density question, and here's what I got.

Say we start with a standard structure and standard armor Commando and Atlas.

(1) 10% of the weight goes to structure, and 20% (approx) goes to armor. The structure and armor should be the same density, given how it's all swappable (in case of structure as scrap). At this point, we've used up 30% of the total weight and at the same density for both mechs.

(2) The remaining 70% of the tonnage goes into the 53 free slots. We know the slots are the same size, since if they scaled with tonnage a 3 slot DHS on a Commando would take up less than 1 slot on an Atlas, which is not the case. It doesn't matter which slots you fill and which slots you leave empty, you have 17.5 tons versus 70 tons going into the same sized 53 slots.

We don't know what % of the total volume the structure and armor take up, but we do know that for 30% of the total tonnage, the density is the same, and for the other 70%, the Atlas is much more dense (4x more). Right? Or can there be other 'empty space' in a mech that is not part of the slot system?

[Y E O N N E]

Y'know, I always rewrote the canon in my head and declared a 'Mech's tonnage to be its carrying capacity rather than total mass.

[InspectorG]

Yeonne Greene, on 04 March 2018 - 08:43 PM, said:

And then there is the Nightstar, which is enormous for 95 tons and makes the Banshee look lithe.

Hey, i cant say PGI is consistent aside from nerfs and 3 missing pillars.

[Nightbird]

Yeonne Greene, on 05 March 2018 - 08:15 PM, said:

Y'know, I always rewrote the canon in my head and declared a 'Mech's tonnage to be its carrying capacity rather than total mass.

That could be it... if they didn't preserve 5 or 10% as structure it would be obvious actually.

In my mind, I had this notion of super strong and light weight materials. Basically both structure and armor would have the density and weight of today's foam but be capable of enduring enormous pressure and stopping cannon rounds. Basically the next step forward from ceramic armor and carbon fiber structure.

[Nicodemus Rosse]

Getting scaling right would involve taking into account things like:

• Armor is very, very, very heavy for its volume. FF armor is lighter by volume, so the same weight is much bulkier, but for some reason the extra volume is inside the 'Mech and takes up critical space—presumably because each 'Mech chassis has joints designed around a certain exterior volume, and bulking the whole thing up like an overinflated robot balloon would cause it to hit itself when moving. So no change to overall volume.
  
• Structure is moderately dense for its volume, but remember that the internal structure includes actuator weight, and myomers are less dense than the skeleton is, so even with a really dense skeleton, all the actuators and myomers mean that the structure is "moderately dense" on average. ES structure is foamed so it's bulkier but lighter; that doesn't change the external volume, though, because it crowds out internal space (and thus uses crits). So no change to overall volume.
  
• Ballistic ammo (esp. the depleted uranium penetrator rounds the novels usually describe autocannons firing) is really heavy for its volume. So are batteries used to accumulate energy weapon charges from the engine between shots, and the capacitors of gauss rifles.
  
• The different engine ratings are all roughly the same (small) volume because the majority of the engine's volume is the shielding, not the power plant core, so the increase in weight comes with very little change in volume.
  
• Some 'Mechs have empty space inside when loaded with smaller, denser heavy equipment as compared to the max volume the chassis can equip; you'd need to set the base volume of each chassis based on the bulkiest loadout of any possible variant, and one with denser equipment would just have more open internal spaces in the weapon bays, etc.

Lots of variables apply as well. For example:

WWII German Maus tank (1944): ~135 m³, weighs ~188 tons, ~1.4 tons/m³
Modern M1A2 Abrams tank (1986): ~70 m³, weighs ~65 tons, ~0.92 tons/m³

Made roughly 40 years apart, but the newer one is much less dense, much faster, and at least as powerful. So older, lower-tech 'Mech designs (i.e. anything designed pre-Star League or during the Succession Wars, or in the periphery) would plausibly be bulkier by weight, and Clan designs should by all rights be more compact by weight, so even any two 75-ton 'Mechs might have different volumes and still maintain the same operating weight.

Much as I'd love to have perfectly logical, mathematically sound scaling, it's by necessity more art than science in this case.

[LordNothing]

i figure its one of those things where by the size of the measuring stick as a large impact on the result when measuring the length of a shoreline. also the problem with using volume is that you should have used surface area instead. in theory surface area should be proportional to volume but that doesnt mean that the proportion is the same between different mechs even of similar tonnage. this combined with the thing at the beginning result in wierd mech sizes. that said i dont think its a big deal.

Edited by LordNothing, 05 March 2018 - 09:35 PM.

[Nightbird]

Nightbird, on 05 March 2018 - 08:07 PM, said:

So I got a question. I've been thinking though the density question, and here's what I got.

Say we start with a standard structure and standard armor Commando and Atlas.

(1) 10% of the weight goes to structure, and 20% (approx) goes to armor. The structure and armor should be the same density, given how it's all swappable (in case of structure as scrap). At this point, we've used up 30% of the total weight and at the same density for both mechs.

(2) The remaining 70% of the tonnage goes into the 53 free slots. We know the slots are the same size, since if they scaled with tonnage a 3 slot DHS on a Commando would take up less than 1 slot on an Atlas, which is not the case. It doesn't matter which slots you fill and which slots you leave empty, you have 17.5 tons versus 70 tons going into the same sized 53 slots.

We don't know what % of the total volume the structure and armor take up, but we do know that for 30% of the total tonnage, the density is the same, and for the other 70%, the Atlas is much more dense (4x more). Right? Or can there be other 'empty space' in a mech that is not part of the slot system?

Still hoping for some insights on this, or if anyone has a source on mech tonnage being only carrying capacity, that would also work!

[MW Waldorf Statler]

the right arm of the Centurion has the same dimension and Proportions of the left arm =same density ? the Right arm have a Big Canon, the left arm nothing ...in "reality" no mechlab no ammo to put in the Legs and come magical to the left torso ..all mechs build around her specific Engine ,Weapons ...

what do I do to save weight ... I provide parts with recesses, indentations, holes and neuter inner structurs or bulges to increase their stability with thin material ... of course not visible in the outline but on the surface or by volume measurement which does not funtionalwith compleced bodies

Edited by Old MW4 Ranger, 27 February 2019 - 03:33 AM.

[Magnus Santini]

Samial, on 04 March 2018 - 04:09 AM, said:

Man what i'd give for an Atlas the size of a Commando.

Purusee, on 04 March 2018 - 02:13 AM, said:

Scratch the Mini-Atlas, I want an Assault-Commando!

If they give us this, I am ready for them to bring back knockdown. Just leave your mechs in the drop zone and it will be Bowling for C-Bills.

[Moonlight Grimoire]

As someone who has been working on getting every mech in MWO into a 3d printable state (it is not as easy as it sounds, PGI took some nasty short cuts in making mechs) I will say this, there are some true liberties taken with volumetric scaling. Now I get that "oh not everything will have the same density" stuff, but, strip off all the equipment down to just the bare bones of the mech, the structure, the endosteel structure, all IS should follow a trend line and all clan should follow another (because clan's figured out how to make it better by being less dense but just as strong, hence half slots for the same weight savings). This means that the skeleton of mechs is the only real mass to go off of for our scaling, the rest after all is weight the skeleton can hold safely. From even this we get the same sliding scale and the same issues in the end. I mean I can see arguing have a degree of variance from one species to the next but all variants are the same to keep from true uniformity, but what PGI did is beyond that.

[Nightbird]

I checked out the models in your sig, wow, very nice!

[InvictusLee]

Baby Atlas, and Mini Knight Skins for bolt-ons please!!!

[Davegt27]

some of us built plastic models when we where kids

like 1/72 scale F15

maybe volumetric scaling is not the best way to approach the scale problem


on another note PGI already worked on the Mech scaling and did what they saw fit
why revisit the problem

[Nightbird]

Smaller assaults with geometry at the size of today's heavies will allow damage to spread like today's heavies? You may have noticed the excessive amount of quirks needed to maintain today's scale. (The heavies with the armor of assaults, the assaults with the armor of 125 ton mechs?) The result of a proper scale is increased TTK w/o all these quirks, which PGI has been aiming for, and some have been complaining about, for all mechs except 20-25 tonners if those are used as the basis.

Finally... PGI can't do all these re-scales manually... they need to write an app that will automate this... then it'll take 1 minute for each chassis. But but but... PGI can't write an App it's too hard? Bullcrap... as the first post is intended to show, just take the current model in plain text data, and multiply all (x,y,z) coordinate values by X%. All the app needs is a text parser and math operator, you can do it in any number of languages in 1 day.

Edit: ok, when I wrote the above I just woke up to a storm outside and my battery backup beeping

Edited by Nightbird, 07 March 2018 - 07:56 AM.

[Y E O N N E]

The app has tobe smarter than that. Part of the problem with using volume and surface area is that both are going to count greeble and functionally neutral dead space. That will give you some 'Mechs being too big or too small for their weight relative to a 'Mech that seems "right" .

[Fox the Apprentice]

Yeonne Greene, on 07 March 2018 - 11:32 AM, said:

The app has tobe smarter than that. Part of the problem with using volume and surface area is that both are going to count greeble and functionally neutral dead space. That will give you some 'Mechs being too big or too small for their weight relative to a 'Mech that seems "right" .

While you are correct, it might still be better than what we have currently. Or it might not. We'll never really know.

[Y E O N N E]

Fox the Apprentice, on 07 March 2018 - 12:14 PM, said:

While you are correct, it might still be better than what we have currently. Or it might not. We'll never really know.

It's already partly to blame for what we have now. PGI may not have used a single constant to root volumetric scale across all weights, but they still had a constant for each weight class. Within each set, you have problem 'Mechs for precisely the reason I mentioned.

[Nightbird]

Yeonne Greene, on 07 March 2018 - 11:32 AM, said:

The app has tobe smarter than that. Part of the problem with using volume and surface area is that both are going to count greeble and functionally neutral dead space. That will give you some 'Mechs being too big or too small for their weight relative to a 'Mech that seems "right" .

Well, I would still give the human the job of determining what % The benefit of the app is simply the ability to say 'hey! this mech is clearly too large!' and have PGI be able to say, 'oh! we agree! we'll spend 30 seconds to resize it down 5% more'. Right now, the process is just too long and tedious as the models for the mech, weapons, and paint job have to be manually redone...

Edited by Nightbird, 07 March 2018 - 12:43 PM.

[Grus]

HEY YOU LEAVE MY COMMANDO ALONE! lol