Yes 'Exoskeleton'?

[cvalue13]

A thought experiment to set this up:

Take the CT "endoskeleton" as we know it based on recent photos. Then consider two different vehicles built around it.

Both have identical suspension robustness.

In the first vehicle, hang upon it conventional sheet metal like the Model Y, fastened to the frame/casting in the same manner as a Model Y. Call this first vehicle the 'Body In White' version of the CT.

In the second vehicle, hang upon it instead 3MM SS, and rather than fastening it conventionally, use a multitude of fasteners, adhesives, brackets, etc., to effectively 'weld' the SS to the internal castings/frame, including across the joints of the casting/frame seams. Call this second vehicle the 'Exoskelton' version of the CT.

Now consider the operational/functional delta between the Body in White CT and the Exoskelton CT.

Cory at Monro appears to be indirectly maintaining that those two vehicles will have identical payload/towing capacities, and identical behaviors when put under identical tortional stress. Doesn't that seem a bit precarious?

From my armchair, I suggest it's possible that Tesla both:

• manages to make the 'Exoskeleton' CT materially outperform the 'Body in White' CT, and
• given that material difference in capability, all owing to the outer skin construction, can justifiably claim that the performance of this CT is due to its "exoskeleton"

More on this and the 'other' type of exoskeleton:

Sponsored

 

[cvalue13]

 

[cvalue13]

All that said:

• if Cory’s assumptions that the SS panels are hung just as on the Model Y are correct, then I’d tend to side with him about the CT not having operational structure from the panels (though the doors, hood, and tailgate are each independent, self-contained exoskeletons themselves)

• no matter how strong the panels may be, the CT is only as capable as it’s weakest link; as Cory discusses, heavy haulers tend to have solid axels - esp in rear - for a reason (rather than independent suspension), and tend to have mechanical (rather than air) suspension.

• finally, on another view, even if the panels add operational structure, it’s possible they do so only to make up for the shortcomings of the “endoskeleton” - that is, if the casting/frame joints and structure aren’t terribly capable alone, the panels could be working not so much to ‘amplify’ capability, but instead just to achieve acceptable performance

Afterall, there are 1/2 ton trucks that, if optioned correctly, can achieve materially the same operational capabilities as the CT’s claimed specs. Viewed that way, the strengthening panels wouldn’t be using the exoskeleton to offer previously unknown capability, so much as to allow the Model Y-like construction techniques the boost needed to compete in the first place. Less of a revolution in capabilities, and more of an alternative method to achieve the ~same outcomes.

Less being “actually tough” and instead being “also tough.”

On that view, the novel approach to construction boils down more to the advantages of manufacturing costs for Tesla, and scratch/dent (bullet?) resistance for customers. Not for nothing, as those are good outcomes.

But it would take the sheen off the marketed angle of the exoskeleton. Especially for people who really use their trucks, as opposed to urban posers (like me) - as the “real” truck guys tend to not care as much about scratches/dents (and are doing the shooting first).

 

[JBee]

All that said:

• if Cory’s assumptions that the SS panels are hung just as on the Model Y are correct, then I’d tend to side with him about the CT not having operational structure from the panels (though the doors, hood, and tailgate are each independent, self-contained exoskeletons themselves)

• no matter how strong the panels may be, the CT is only as capable as it’s weakest link; as Cory discusses, heavy haulers tend to have solid axels - esp in rear - for a reason (rather than independent suspension), and tend to have mechanical (rather than air) suspension.

• finally, on another view, even if the panels add operational structure, it’s possible they do so only to make up for the shortcomings of the “endoskeleton” - that is, if the casting/frame joints and structure aren’t terribly capable alone, the panels could be working not so much to ‘amplify’ capability, but instead just to achieve acceptable performance

Afterall, there are 1/2 ton trucks that, if optioned correctly, can achieve materially the same operational capabilities as the CT’s claimed specs. Viewed that way, the strengthening panels wouldn’t be using the exoskeleton to offer previously unknown capability, so much as to allow the Model Y-like construction techniques the boost needed to compete in the first place. Less of a revolution in capabilities, and more of an alternative method to achieve the ~same outcomes.

Less being “actually tough” and instead being “also tough.”

On that view, the novel approach to construction boils down more to the advantages of manufacturing costs for Tesla, and scratch/dent (bullet?) resistance for customers. Not for nothing, as those are good outcomes.

But it would take the sheen off the marketed angle of the exoskeleton. Especially for people who really use their trucks, as opposed to urban posers (like me) - as the “real” truck guys tend to not care as much about scratches/dents (and are doing the shooting first).

First I'd like tot refer you to a post I made on the "No Exoskeleton" thread in response to similar queries you had there: (I can copy it here if you'd like)

https://www.cybertruckownersclub.com/forum/threads/no-exoskeleton.8357/post-148070

Thanks again for taking your time for making these posts with pictures. The effort is a welcome addition to the discussions here, and highlights some more methodical ways of coming to some design conclusions around the CT.

In saying that, some of you assumptions are correct, some conclusions as well, but at a base level I think it's important to understand that the way load paths work is that they go the path of least resistance, typically in a vehicle that means in fairly straight lines in accordance to where the force is applied. Even after doing years of FEA simulations you have to be careful at what point your attach your force/load vectors and ground points, to make sure that they match how they are in reality.

For example: constraining a lever in such a way that you have a load at one end, a fulcrum in the middle and a force at the other would demonstrate how a simple lever works. But, having a fulcrum at one end, a load at the other and a force in the middle would be better representative of a suspension arm lever, but would be fundamentally different to how the forces would react mechanically with eachother.

Let have a look at some of your diagrams to give some more insight:

In reality those torsional forces are not applied to the outside corners of the vehicle rather they create forces that compress the air spring into the top of the cast and primarily into the cabin frame, both in the front and the diagonally opposed rear as well.

The cabin frame, consisting of structural glass and a structural battery pack and the cabin roll cage frame all act together to lift the vehicle by just two wheels, if the other two opposing wheels leave the ground.

Now if I attach a front fender that is permanently attached for structural exoskeleton purposes, the part of the fender that is in front of the front suspension air spring can be essentially neglected, as no load is presented to it by the airspring, because there is no other element in the front that a force can be applied too. It is in the air. The same is the case for the rear fender. However, going towards the rear the upper cabin frame and glass is compressed and the lower frame and battery is put under tension, with loads going through opposite sides of a diamond/square shape(as seen from above) and truss (as seen from the side).

In your diagrams above you have clearly demonstrated that the casts sit on the airsprings, and even in the case of the rear airspring the substantial cast there put compression load to the top of the cabin frame, while the lower wheal arch at the bottom pulls tension through the structural pack. Now for the fender to do anything structurally, it would only do so between the airspring to the outside of the cabin frame. However, in this example the load path is diagonally through the vehicle to the opposite corner wheel, and as such the tension forces travel through the structural pack directly, and through the roof structure and glass. To go through the skin outside of the cast would be a major detour for these loads.

In considering this we also have to consider from what source the load on the structure is created. For example the structural pack is not only a structure, but is the single heaviest element in the vehicle construction. So in this case the load of the battery on the airsprings originates on one side of the pack to one airspring, and on the other side of the pack to the other airspring. They in no way have a way to interact with the fender skins.

We can take this further, by placing a 1 ton load into the bed. In this case the load in the bed would go straight down and into the rear cast, airpsring suspension and into the wheel. If stationary, and the load centered over the rear axle, no static load would be transferred to the front wheel that originated from the payload in the bed. Further, pending the weight balance front to rear on the CT, it's highly likely that the payload will force both rear wheels to the ground, and only one front wheel would be in the air.

Now structurally, the payload capacity of the CT, including the passengers and frunk area are not that much better than a F250 which rides a conventional ladder frame. Accordingly, there is nothing inherently necessitating an additional structure in the CT design beyond the body in white as shown. Even if the load path were to make its way through the fender skins, it would actually cause potentials problems with the skin by making it bend or warp under the pressure. It also means that repairs to the four most common impact points, in light crashes would be much more complicated and expensive than would be necessary.

If you like we can go through each scenario one by one to see if we can find one where a structural skin would be beneficial, but as yet I have to find one, given that we now know the nature of it's internal structure.

 

[Coolbreeze704]

Got it. Makes sense.

 

[cvalue13]

Thanks for all the thoughts and additions, a few queries:

In reality those torsional forces are not applied to the outside corners of the vehicle rather they create forces that compress the air spring into the top of the cast and primarily into the cabin frame, both in the front and the diagonally opposed rear as well.

Yes, I appreciate all this, but take it a step further

so the force is in the cabin frame, as you say.

from there we seem to be miss communicating.

in a Model Y, that force on cabin frame then strains, e.g., the weldments at the joints between the casting and the central cab. Too much force, and that joint flexes/deforms/separates. No? (This of course assuming the suspension doesn’t absorb first.)

If so, then the conjecture of the post is that this same joint between casting/cab has not only the “normal” weldments seen in a Model Y, but also additional “weldments” created by using the SS panel essentially as a mending bracket across both the cab/casting.

Accordingly, the force to the cabin frame you point out is now supported not only by the “normal” weldments there (which you’d agree are part of the cabin frame taking the force), but now also the SS panels - and I don’t understand why these weldments you’re excluding as part of the cab/casting assembly. .

In effect, if the 3MM SS QPs are welded across the cab/casting seams, they become part of the cab/casting structure, just the same as the “normal” weldments in a Model Y between cab/casting.

But because I think you’ve gathered all this from my post, it feels like somewhere we’re still talking past one-another.

Put differently, given the conjecture I’ve laid out, I don’t understand how you’re still focused on explaining that the suspension transfers force to the cab/cabin - this much is clear, and we agree on.

The remaining question is what all is “included” in the cab/casting structure. In the Model Y, the weldments between cab/are included. In the CT, if the panels are similarly “welded” across, why aren’t you similarly including them in the cab/casting taking the force?

 

[cvalue13]

Accordingly, there is nothing inherently necessitating additional structure in the CT design beyond the body in white as shown.

second, the above isn’t necessarily true.

set aside for a moment the issue of suspension robustness.

past that, an F250 has a frame capable of withstanding the operational forces.

with the CT, they’ve used the different approach of an aluminum casting set plus a steel cab. it’s an open question as to whether those structures collectively are more or less ridged than the F250 frame.

if alone the casting/can are less ridged than the F250 frame, then additional support is required to achieve the same stiffness as the F250

we have no idea what the cab/casting frame capabilities are, alone. could be more or less than the capabilities of the F250 frame.

 

[cvalue13]

If you like we can go through each scenario one by one to see if we can find one where a structural skin would be beneficial, but as yet I have to find one, given that we now know the nature of it's internal structure.

one last shot at a reframing:

FRAME 1:

take one known casting/cab frame - call it “Frame 1”​

use an approach to weldments between casting/cab, call that approach’s weldments strength “Strength 1”​

apply forces to the frame until the joint between casting/cab flexes/deforms/separates.
Call the turning point of flex etc “Force 1”​

FRAME 2:

now take a second casting/cab frame. Call it Frame 2

Except in Frame 2, at the joints, use a a methodology of weldments that is 2X as strong. Call this “Strength 2”​

Now, apply Force 1 to Frame 2​

Your posts appear to suggest that despite the 2X strength of the weldments in Frame 2, when Force 1 is applied it will flex/deform etc identically

that makes no sense to me

 

[JBee]

Thanks for all the thoughts and additions, a few queries:



Yes, I appreciate all this, but take it a step further

so the force is in the cabin frame, as you say.

from there we seem to be miss communicating.

in a Model Y, that force on cabin frame then strains, e.g., the weldments at the joints between the casting and the central cab. Too much force, and that joint flexes/deforms/separates. No? (This of course assuming the suspension doesn’t absorb first.)

If so, then the conjecture of the post is that this same joint between casting/cab has not only the “normal” weldments seen in a Model Y, but also additional “weldments” created by using the SS panel essentially as a mending bracket across both the cab/casting.

Accordingly, the force to the cabin frame you point out is now supported not only by the “normal” weldments there (which you’d agree are part of the cabin frame taking the force), but now also the SS panels - and I don’t understand why these weldments you’re excluding as part of the cab/casting assembly. .

In effect, if the 3MM SS QPs are welded across the cab/casting seams, they become part of the cab/casting structure, just the same as the “normal” weldments in a Model Y between cab/casting.

But because I think you’ve gathered all this from my post, it feels like somewhere we’re still talking past one-another.

Put differently, given the conjecture I’ve laid out, I don’t understand how you’re still focused on explaining that the suspension transfers force to the cab/cabin - this much is clear, and we agree on.

The remaining question is what all is “included” in the cab/casting structure. In the Model Y, the weldments between cab/are included. In the CT, if the panels are similarly “welded” across, why aren’t you similarly including them in the cab/casting taking the force?

I think some of it might just be terminology.

By weldments do you mean the interface points between the cast and the cabin frame? I don't think they will weld dissimilar metals at all due to dielectric and differential thermal expansion. I'd expect It's probably a combination of intersecting material, adhesive and fasteners, this would also allow repairability.

If you are concerned about the attachment points, those are considerably more capable than they would first appear. Now if you take a bolted connection, the compression forces are transferred just by the surface interaction between the two parts alone, and the bolt takes no load, and if under tension the bolt would take a load. Now just for reference, a single 12mm bolt has enough tensile strength to support 7tons. That means a few of those attachment points are likely to outperform the rest of the cabin frame.

 

[JBee]

second, the above isn’t necessarily true.

set aside for a moment the issue of suspension robustness.

past that, an F250 has a frame capable of withstanding the operational forces.

with the CT, they’ve used the different approach of an aluminum casting set plus a steel cab. it’s an open question as to whether those structures collectively are more or less ridged than the F250 frame.

if alone the casting/can are less ridged than the F250 frame, then additional support is required to achieve the same stiffness as the F250

we have no idea what the cab/casting frame capabilities are, alone. could be more or less than the capabilities of the F250 frame.

The normal procedure for engineering a structure is to come up with the desired geometry and then calculate using a load analysis the required material properties, like how thick it is. You then add a safety margin, somewhere between x1.5 to x3 is common, to make sure it outperforms the worst case calculation.

If Tesla has done anything similar, which I am sure they did, the skinless structure is likely to be more than rigid enough for any of the various forces it would counter in operation.

This is also another important point: the truss geometry, and high sided sail cast, along with the cabin frame being anchored on top of the front airspring, will result in a significantly higher frame structure with better force distribution. With a space frame design the height between the top and bottom chord separated by a high web element, means that the compression/tension forces are further apart which benefits the structure by allowing less material to be used.

This whole thing would have been optimized to the extreme, in detail that we are not going to be privy to unless we could recreate their exact modeling.

 

[JBee]

one last shot at a reframing:

FRAME 1:

​

take one known casting/cab frame - call it “Frame 1”​

​

use an approach to weldments between casting/cab, call that approach’s weldments strength “Strength 1”​

​

apply forces to the frame until the joint between casting/cab flexes/deforms/separates.​

Call the turning point of flex etc “Force 1”​

FRAME 2:

now take a second casting/cab frame. Call it Frame 2

​

Except in Frame 2, at the joints, use a a methodology of weldments that is 2X as strong. Call this “Strength 2”​

​

Now, apply Force 1 to Frame 2​

Your posts appear to suggest that despite the 2X strength of the weldments in Frame 2, when Force 1 is applied it will flex/deform etc identically

that makes no sense to me

If you refer to my post above I don't think anything will be welded between the cast and cabin. Casts are by there nature difficult to weld to at all.

As per my post also, I don't think the attachment points between the cast and cabin parts are the weakest element, and if they were they would just increase their dimensions to compensate. Overall, I think it's pretty safe to assume that the cabin and casts all together one fairly homogenous structure for transferring expected loads throughout the vehicle.

With modern FEA you basically apply the expected forces to your design geometry and then go hunting for "hot spots" where you might experience material failure. Nowadays with parametric design, the simulation actually runs a few variables to see which variations of the design get the best results. It's actually quite impressive what they can do, and how many digital iterations can be simulated before even building a single part IRL.

 

[cvalue13]

I think some of it might just be terminology.

By weldments do you mean the interface points between the cast and the cabin frame? I don't think they will weld dissimilar metals at all due to dielectric and differential thermal expansion. I'd expect It's probably a combination of intersecting material, adhesive and fasteners, this would also allow repairability.

using “weldments” as a generic term for [take your pick as to how the cab/castings are unitized] in a Model Y

This whole thing would have been optimized to the extreme, in detail that we are not going to be privy to unless we could recreate their exact modeling.

But you’re assuming away the exact point in discussion

what if instead the optimization was made wholistically on the frame/skin combo?

absent that combo, they would have designed a different frame/casting combo

your comment quoted last shows that you are starting from the assumption the frame/casting combo (and it’s weldments) are sufficient on its own

but the point in discussion is: what if it is not - what if the CT was designed to utilize the panels as additional weldments between the frame/casting joints

to be honest, I’m starting to feel that a few years of assumption are impeding your flexibility in understanding the relevant questions at issue

I know the force goes to the casting/frame. But the question is what is included/excluded in the definition of casting/frame. If the panels are weldments of it, they are part of it.

I know Tesla has modeled/optimized for frame integrity. But the question is whether that optimization presumed/included the panels as weldments.

Fact is, you do not know either answer. Not until the final product is deconstructed will we know.

I’m merely pointing out that there exists a scenario where the panels are a bonded portion of casting/cab structure, in which case they are part of the structure that takes the force from the suspension, and for which Tesla has modeled/optimized the design.

 

[JBee]

using “weldments” as a generic term for [take your pick as to how the cab/castings are unitized] in a Model Y




But you’re assuming away the exact point in discussion

what if instead the optimization was made wholistically on the frame/skin combo?

absent that combo, they would have designed a different frame/casting combo

your comment quoted last shows that you are starting from the assumption the frame/casting combo (and it’s weldments) are sufficient on its own

but the point in discussion is: what if it is not - what if the CT was designed to utilize the panels as additional weldments between the frame/casting joints

to be honest, I’m starting to feel that a few years of assumption are impeding your flexibility in understanding the relevant questions at issue

I know the force goes to the casting/frame. But the question is what is included/excluded in the definition of casting/frame. If the panels are weldments of it, they are part of it.

I know Tesla has modeled/optimized for frame integrity. But the question is whether that optimization presumed/included the panels as weldments.

Fact is, you do not know either answer. Not until the final product is deconstructed will we know.

I’m merely pointing out that there exists a scenario where the panels are a bonded portion of casting/cab structure, in which case they are part of the structure that takes the force from the suspension, and for which Tesla has modeled/optimized the design.

So let me get this right: you are asking me to consider a design version where the stainless steel skin "must" take some of the forces in the vehicle structure?

Sure I can do that, I have even been thinking of how I could design a van that can make better use of the skin, like an aircraft fuselage.

But....it is difficult to imagine how this can be achieved given the known geometry of the CT from the photos we have so far. Let me give another example of how I see the panels: Imagine if we welded a 4" by 4" square steel tube section to the front fender, and let it extend past the front bumper by 6ft. Now although the fender skin is fixed structurally to the cast and the cabin, if I don't apply a load to the end of the tube, then what load does the fender have to carry? Just the mass of the tube right? Likewise I cannot imagine where a force can be applied to the fender itself, from the vehicle or payload, that would cause a force to go through a fender, instead of the cast or cabin to the airspring. You could make the fender 10inches thick if you wanted to, but I don't see how those loads can get to the fender in the first place, they'd just take a shortcut through the cast every time. There might be one, but I'd need a specific example to analyze and comment on.

 

[cvalue13]

So let me get this right: you are asking me to consider a design version where the stainless steel skin "must" take some of the forces in the vehicle structure?

On the contrary.

It’s instead that the question being asked is “how do we know that the frame alone performs all structural functionality.”

And you respond “because the frame has been designed to perform all structural functionality.”

Far as I can tell, you’ve yet to state anything that gets at the heart of the matter: a deep engineering reason that it would be impossible or unlikely for an alternative design to exist.

A design where, for example, in a classic Tesla effort to minimize parts and threaded weldments, they either:

(1) minimized these weldments in the casting/cab design, and partially replaced the structural adhesion of casting/cab by use of adhering the SS panels across the casting/frame joint

or

(2) realized that achieve impressive truck functionality, their historical method of adhering casting to frame was insufficient, and so they met the gap by additionally
adhering the SS panels across the casting/frame joint


A critique of either of the above hypos cannot start and end with what amounts to exclaiming “they wouldn’t have to do that if they designed the frame alone to suffice.

That is assuming the conclusion, petitio principii

 

[JBee]

On the contrary.

It’s instead that the question being asked is “how do we know that the frame alone performs all structural functionality.”

And you respond “because the frame has been designed to perform all structural functionality.”

Far as I can tell, you’ve yet to state anything that gets at the heart of the matter: a deep engineering reason that it would be impossible or unlikely for an alternative design to exist.

A design where, for example, in a classic Tesla effort to minimize parts and threaded weldments, they either:

(1) minimized these weldments in the casting/cab design, and partially replaced the structural adhesion of casting/cab by use of adhering the SS panels across the casting/frame joint

or

(2) realized that achieve impressive truck functionality, their historical method of adhering casting to frame was insufficient, and so they met the gap by additionally
adhering the SS panels across the casting/frame joint


A critique of either of the above hypos cannot start and end with what amounts to exclaiming “they wouldn’t have to do that if they designed the frame alone to suffice.

That is assuming the conclusion, petitio principii

Mate, I've been doing nothing other than trying to demonstrate "why" a load path can't go through the skin. You sure it's me and not your understanding how the load paths work, being point to point? Go crack an egg and look.

If you want to get a load to go through the skin from airspring to airsping, we would have to delete the cast and cabin on which the airsprings sit. That's all I would normally need to argue to make the point.

Tell me how a load can be redirected through the skin then, whilst the cast and cabin is in place, because I don't understand the physics of how that works.

Those skins are essentially structurally ornamental and are flapping in the wind until it crashes. To be honest I don't see what is wrong with that, because most of the doors are the same and that's 2/3 of the body surface.

Let's turn the table on the arguement.

Why is it so important, given the other benefits of the SS skin, that they DO structurally improve the body under operational load? They're not even 3% of the mass of the vehicle yet the CT will collapse without them? Then tell me how they do this with a flat sheet with a couple of bends in 3D, whilst the skin is mostly 2D, and how it supports load without warping or twisting from said load, and how it is attached in such a way its repairable but still structural, and how the dimensional accuracy of the CT is maintained without it being attached on the assembly line etc.

In particular though tell me how it carries load whilst being outside of the load path.

Imagine this: one airspring mounting point is the fulcrum, the opposite side airspring is the load, and you have a steel beam that joins them and extends past the load as a lever for say 10ft. Now you use force to lift the load, but you don't do it from the end of the beam 10ft away, rather you do it from only 5ft away from the load. What does the extra 5ft of steel lever behind you help in this situation?

Such is the nature and position of the fenders.

Please explain how it is otherwise.

BTW Just raising a generic question mark on how I make my arguments doesn't invalidate all of the things I say either.

That's why we have sentences and paragraphs, so you can choose your poison and argue a remedy.

Or have you taken to seriously to heart that I recently said there's no-one on the forum to debate anymore? Lol. Sorry I thought my attempt at humour was clear.

Sponsored