LocostUSA.com

Hmm... something wrong there, I can tell just by looking.

From experience, I would say that there is no way that a spaceframe resembling your design could get anywhere near those figures. Even with a fully caged design, you'd struggle to reach the lower figure. And an increase of nearly 17,000lb.ft/degree (that's about 15 times the total stiffness of a book Locost frame!) just by adding the floor? Pull the other one...

The fact that it has jumped so dramatically from your previous numbers (which sound realistic) should have been ringing alarm bells for you.



My design is for a commercial client, so I won't be showing anything of it until the car is launched (which was intended to be at next January's Autosports show, but the programme is slipping dramatically behind schedule at the moment). Even then, I won't be disclosing accurate torsional stiffness figures. Sorry, that's just the way it is.

Ok if you say so but it was confirmed by 2 other engineers (well they got similar numbers), one who is a composites engineer at phantom works

You should drive. It's pretty sweet for a steel frame

You should also read my whole build log. There is some good stuff on there

_________________
My build: http://www.locostusa.com/forums/viewtop ... 18&t=13096


You can build the most awesome thing in the world but at some point, an 80yr old man in a crx is probably going kick your butt on the track... don't ask me how I know.

I skimmed a fair way through. There is, but also some quite scary bits.

Are you sure you're using the right conversion factor from Nm/deg to lb.ft/deg? If I didn't know better, I'd say that you'd got a decimal point in the wrong place, although even then the disparity on the floor figure seems too large (unless you're using a sandwich panel for the floor?)

so if someone is paying you, that means someone thinks you know what you are doing... so why are you on here arguing?? Do you argue with the locost UK guys too?



you should point out the scary bits so that i can fix them!

nope decimal is in the right place and bingo...Sandwich floor carbon/balsa/kevlar.

_________________
My build: http://www.locostusa.com/forums/viewtop ... 18&t=13096


You can build the most awesome thing in the world but at some point, an 80yr old man in a crx is probably going kick your butt on the track... don't ask me how I know.

sam_68,

(edited)

I argue with everyone - even people who are paying me - when I think they're wrong.


Sandwich panel floor might account for a 55% difference in stiffness, though even that seems way too large, particularly since you don't appear to have any tubes triangulating down from the two upper curved main rails to the tubes along the edge of the floorpan, so it's relatively poorly tied in to the structure as a whole.



But I'd put money on there being something wrong with that analysis, somewhere. The figure is just way too high for such a poorly triangulated frame (and on experience of the performance of Ar-i-el's very similar At-om chassis).

You appear to be a pretty close copy of the At-om in most respects, so it might be informative to reflect that the prototype At-om chassis had an FEA calculated torsional stiffness of 1,352Nm/deg (997lb.ft/deg) for a weight of 47 kilos, and the fully FEA-developed chassis ended up with 6,030Nm/deg. (4,447lb.ft/degree) - without sandwich panel floor but with much larger diameter main tubes than you appear to have - for a weight of 127kg. The final production chassis has been verified by testing, and has subsequently been subject to further analysis in the form of the Niche Vehicle Network-funded project to fabricate a titanium version.

You can see why I'm more than slightly skeptical, when faced with claims of 21,141lb.ft/degree (nearly 5 times greater!) stiffness with only 67% of the weight, on a chassis that appears to be almost a straight copy?

Your original figures of 2,500lb.ft/degree for a structure weight of 220lbs were much more believable and pretty much on the money for a comparison against the known performance of the At-om, given the smaller diameter main tubes.

The chassis pic is showing the "optimized" version so it shows the tubes at 1" instead of 2" which the "initial" or actual is based on.

only thing similar between mine and the At-om chassis is the main tube look. My terminations and such are very different.

Its all triangulated. Where the At-om falls short is the tub. There is no "floor frame" and the tub is held on with a few screws which adds very little torsional rigidity if at all. here are some pics that show my floor structure. It makes a very big difference.

_________________
My build: http://www.locostusa.com/forums/viewtop ... 18&t=13096


You can build the most awesome thing in the world but at some point, an 80yr old man in a crx is probably going kick your butt on the track... don't ask me how I know.

I disagree (the scuttle/front bay arrangement appears to be a virtual copy) and no it's not - there is absolutely no triangulation between the mid-height curved rail and the rail along the edge of the floorplan, for a start, but I'm seeing lots of untriangulated rectangles on that chassis, so I don't know how you can even think of making that assertion!

There's nothing on those photos that changes my view that the torsion figure you're quoting is seriously erroneous.

But we're drifting way off-topic. If you want to discuss the design of your own chassis, the better place to do it would be on your build thread?

What?! Never mind... i am not getting pulled in more. I am officially over all of this nonsense.

just keep throwing out your knowledge on carbon, etc and disregard everything else, give yourself a pat on the back and call it a day.

_________________
My build: http://www.locostusa.com/forums/viewtop ... 18&t=13096


You can build the most awesome thing in the world but at some point, an 80yr old man in a crx is probably going kick your butt on the track... don't ask me how I know.

Well, I think it would be interesting for the forum as a whole to discover how you have managed to achieve a stiffness:weight ratio 6 times better than a professionally optimized exoskeleton
of very similar design.

Personally I think we're more likely to see flying pigs - possibly with the assistance of jetpacks and unidirectional carbon fibre wings - but there you go.

So if one were using 2 layers of standard weave cloth, you could use two layers of unidirectional cloth at 90 degrees or perhaps better some number like 70 degrees or whatever you thought was optimal. You can even buy fabric that already consists of multiple layers of unidirectional fibers. I think I have some with 3 layers.

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.

On a practical as opposed to purely theoretical note.

My rough chassis design is basically three boxes and a floor when looked at from front to back.
Two sill boxes and the center tunnel.
Center tunnel can be most any shape I like, thinking angled sides and round top.
I will need to accommodate a shift lever and cable or linkage through the center box.
Side sill boxes flat on the bottom and outside, angle on the inside, flat topped.
Narrow chassis so the inside angles will also form the seat bases.
At the rear bulkhead I can form another box, two if I use a sandwich form at the bulkhead and another angle intersecting the inside sill to form the lower back of the seat.
I really need to learn 3D CAD for illustration, I tend to do a lot of building with no more than a pencil sketch.
Rear will have to take up a sub-frame for the drive-train.

If I go air-cooled engine or rear radiator I will not need to route anything through the sills.
In that case I'm not sure if I should add a round tube, intersecting flat support, or maybe expanding foam?
If I use a water-cooled front radiator engine I may have to route water pipes through the sills so expect that any inside support ends up being round tube.

Intent is to make this light enough that even a lowly Fiat 850 or Sunbeam Imp drive-train would be entertaining and a bike engine alarming.
Original design was VW Bug based, steel tube chassis and fairly heavy fiberglass body, it still went pretty well.

Where this will be most complex is going to be the forward scuttle as I have to put my feet through it and also attach a space frame to hold the front suspension.
No real forward firewall, feet go strait out front as this was essentially a two-place Formula Vee for the street.

My one advantage is that the chassis will be topped by a body that should go a fair ways toward closing the top of the box.
I would bond it on but that would make servicing too difficult.
So it will have to be lots of inserts and screws.

I think I'm going to my PC for a while and try to get my head around the 3D CAD.

_________________
If I must be a one-man PC free zone, so be it!

Yes, you could, but you're then relying entirely on the inter-layer adhesion of the resin matrix, rather than the fibres.

Place your hands together palm to palm, then rotate them back and forth against each other through 45 degrees: this is the action you're trying to resist within the material when the forces on the panel are in shear parallel to the panel's surface. The interface between your palms is equivalent to the interface between your separate layers of fabric. Of course there's a film of resin matrix bonding them together, but hopefully you'll appreciate that it would be a lot stronger if there were interwoven fibres crossing back and forth across that interface?

We have thousands of yards of the stitched uni on a -45/45° that was used for aircraft wings and other stuff. They are held together with a few strands of glass. I haven't figured out what to do with the remaining after Boeing lost the contract.

"Woven fabrics are strong reinforcements because the fibers are bundled into yarns oriented in just two directions. The warp and fill yarns run at 0 and 90 degrees respectively. Thus, fabrics are anisotropic, or strong in only two directions. Fabrics need to be oriented so the fiber yarns run parallel to the expected loads. If extra strength is needed in a different direction, another ply must be added at an angle to the first. The most common angles are +/- 45 degrees."



http://www.formula1-dictionary.net/Big/ ... Racing.pdf

"The twill is nice and all but it introduces voids in the cloth which must be filled in with epoxy resin to give it any strength. This means that part of the strength of the carbon sheet lies in the strength of the epoxy. Epoxy is strong, but heavy. Having more carbon in the sheet is better. Therefore, using a unidirectional layup where the fibers are all oriented in one direction with no weaving gives that layer more carbon than a woven layer."

Just a bit more information (I was a guest speaker here a few years ago, in private/military)
https://www.researchgate.net/publicatio ... OXY_MATRIX

I was also a speaker here last year (in private, military nature) This year its Lisbon again the 6th-13th discussing advanced composite reinforce plastics but its not a conference. Customer stuff.
http://www.lamp-project.eu/events/iccs1 ... structures

_________________
My build: http://www.locostusa.com/forums/viewtop ... 18&t=13096


You can build the most awesome thing in the world but at some point, an 80yr old man in a crx is probably going kick your butt on the track... don't ask me how I know.

Indeed; which is why (as it goes on to say) you need more than one layer even of a balanced weave fabric, with 45 degree orientation between the layers.

But the obvious corollary is that since unidirectional fibres are strong in only one direction, you need twice as many layers if you want the composite to be equally strong/stiff in all directions.

In fact (and equally obviously) you'll need a minimum of 4 unidirectional layers vs. 2 for balanced, bidirectional fabric, before it becomes reasonably isotropic.