¯\_(ツ)_/¯ said:All of the Alumalite I can find is extremely thin- both at the core and the inner and outer sheet faces. 1/4" of honeycomb with .016" Aluminum on both sides is probably not up to what we're asking for in a chassis.
What if you cover the whole (final assembled) thing in a layer or two of fiberglass/carbon fiber as well?
In reply to Robbie :
Maybe, but at that point why not just do a cut and fold with the carbon and some sort of core instead? Then you don't have to worry about figuring out how to get your composites to adhere to a nonuniform aluminum surface.
re: Alumalite
the "best" info I could find was a pdf linked from here... which indicated that they didn't want a stack of more than 100 of them on a pallet, since that'd be ~3,000lbs.
looking at it, the core is oriented similar to cardboard, where it runs "longitudinally" instead of "vertically"... so it's more likely to compress with a strong enough force applied perpendicular to the face. off hand math says 0.75lb/in^2... although, that seems low
i.e. yeah, it's probably too weak by itself... and would be better for a self-jigging semi-monocoque design with square-tube aluminum reinforcement... but that'd take some figuring.
Please excuse my ignorance on this. Does an autoclave inherently have a heating function or for the purposes we're discussing is a vacuum chamber sufficient for curing carbon fiber or fiberglass?
If it does have a heating function, what sort of temperatures are we talking about? Would it be practical to build an autoclave for something of this size or could a powder coating booth possibly be used instead?
In reply to stroker :
yes, an autoclave in inherently heating and compressing.
iirc, generally you need an autoclave if you're using pre-preg composites. If you're doing wet-layups, then you can rely on vacuum bagging. The main difference between the two is that pre-preg has "better proportioned" amounts of epoxy, while a wet-layup will most likely have extra epoxy in the finish part so that you don't have "voids" that will create a less-strong part (although, there's still a chance that that might happen).
dunno if a powder coating booth is possible... but there are DIY Carbon Fiber autoclave articles out there.
sleepyhead said:In reply to stroker :
but there are DIY Carbon Fiber autoclave articles out there.
I was able to find this which seems to be relevant to our interests.
I just wanted to add to this thread because composite structures and honeycomb sheets were part of my professional life a few years ago. Things are constantly evolving in that field, so things might have changed, but some core basic problems remain.
Structural composites construction (especially CF) come with a few major hurdles for a grassroots builder.
First is the directional aspect of composite construction. You have to know the constraint path through your parts. That's easy enough for simple mechanical parts like a shaft or disk brake. But for a complex structure, that means precise FEA analysis. Not something for the DIYer. And simply adding another layer isn't a good solution because if the stress is perpendicular to the strains, that other layer does absolutely nothing. You have some vendors that will tell you that their sheets are "quasi-isotropic" (taken from that dragon plate website on page 3), but that's complete BS. In testing, if you end up with a stress that's at an angle just between two major strain direction, you end up with cracks. Carbon is very brittle.
The second problems comes from the construction of the parts themselves. A lot of pre-made composite parts under-perform constantly in tests because of an excess of adhesives. Basically, it means, that the builders use too much epoxy between the fabric. And the strength goes down really quickly when you have too much epoxy. So a lot of parts are closer to a carbon reinforced plastic parts than an actual CF composite construction and will be outperformed by a basic 6061-T6 part. IMHO, if you want a CF structure that will perform, you need some pre-impregnated fabric laid down by pros and then cured in an autoclave. I have never seen good results without that.
What I think is the real biggest problem of composite construction is quality control. For those interested, there is a really good explanation of the problems of composite construction that the Boeing team faced during the development of the X-32 in this documentary. Even with experts and pros, they had to scrap the first wing they did because of voids founds in the quality inspection. And I think this is the reason why composite construction hasn't really taken off yet. Every part has to be inspected individually. That's expensive, time consuming, requires expensive tools and you have to restart from the beginning if a problem is found. When I was doing testing, we saw a lot of problems (voids, excess epoxy and misaligned fabric) and under-performing parts.
That problem is also present in honeycomb panels. You have panels that either have voids where the skin isn't glued entirely to the honeycomb, or a panel that is overly heavy because they drenched it adhesive.
That's why I think that for an amateurish build, you should go either with a good old steel tubular spaceframe or with an Elise-esque chassis:
In reply to fanfoy :
The Elise is bonded aluminum? Are the adhesives for that sort of thing commercially available? Is there specific clamping forces needed? Bonded aluminum always seemed like a good idea, especially with the advent of large laser cutting tables.
from memory, the Elise chassis is bonded, and strategically riveted... the rivets provide the clamping forces, and add a fastener to keep the bond from "unzippering"
mazdeuce - Seth said:In reply to fanfoy :
The Elise is bonded aluminum? Are the adhesives for that sort of thing commercially available? Is there specific clamping forces needed? Bonded aluminum always seemed like a good idea, especially with the advent of large laser cutting tables.
Yes, bonded extruded aluminum parts. The system works well, but is very sensitive to heat. Which is good for a mid engined car, but sucks for front engine- so the Elise was a great lightweight car, whereas the Aston Martin Vantage (made by the exact same methods) was super heavy. I remember seeing some data where the failure temp of the adhesive was way below the normal hot weather soak temps that we saw in the DB7.
alfadriver said:mazdeuce - Seth said:In reply to fanfoy :
The Elise is bonded aluminum? Are the adhesives for that sort of thing commercially available? Is there specific clamping forces needed? Bonded aluminum always seemed like a good idea, especially with the advent of large laser cutting tables.
Yes, bonded extruded aluminum parts. The system works well, but is very sensitive to heat. Which is good for a mid engined car, but sucks for front engine- so the Elise was a great lightweight car, whereas the Aston Martin Vantage (made by the exact same methods) was super heavy. I remember seeing some data where the failure temp of the adhesive was way below the normal hot weather soak temps that we saw in the DB7.
There are a bunch of adhesive that are commercially available and it's a field that is expending very rapidly. With the F-150 now made of aluminum, I think we are about to see the prices drop really fast and the technology go up. Clamping force is usually needed, and that's why Lotus riveted the panels together. It also helped with crash tests if I remember correctly.
Edit: And yeah the first adhesive that Lotus used was very sensitive to heat but they changed a few years later to address that issue.
Though these were interesting and relevant.
steel tube frame with carbon panels bonded to it:
http://www.radonsport.com/RaceTech151-Radon.pdf
Peter Elleray on chassis construction:
Jah29 said:Though these were interesting and relevant.
steel tube frame with carbon panels bonded to it:
http://www.radonsport.com/RaceTech151-Radon.pdf
Peter Elleray on chassis construction:
Thank you for posting that. I didn't know about the Radon. Now that is some brilliant engineering that could be used by grassroots people.
In reply to Jah29 :
Thanks for chiming in about that.
Some take aways from me...
1) were the mention of Dyneema in the same breath as Kevlar. I'd been pondering it's use as a Fiberglass replacement for wing structures, and clearly I need to do some more reading up on it, w.r.t. mechanical properties. Certainly, it would seem to be more "friendly" to use than Kevlar, based on it's current use in ultralight outdoors products.
2) This is the second reference to the use of OpenFoam for CFD... which is off-topic. but clearly I need to go out and dig into that at some point.
3) I dig the "hybrid" setup... especially if/when 3D printers are able to produce the "interface" widgets (?) to pull tubes together and interface them with bulkheads and/or panels.
actually, I wonder if about the ability to use ABS as a core material?
fanfoy said:I just wanted to add to this thread because composite structures and honeycomb sheets were part of my professional life a few years ago. Things are constantly evolving in that field, so things might have changed, but some core basic problems remain.
Structural composites construction (especially CF) come with a few major hurdles for a grassroots builder.
First is the directional aspect of composite construction. You have to know the constraint path through your parts. That's easy enough for simple mechanical parts like a shaft or disk brake. But for a complex structure, that means precise FEA analysis. Not something for the DIYer. And simply adding another layer isn't a good solution because if the stress is perpendicular to the strains, that other layer does absolutely nothing. You have some vendors that will tell you that their sheets are "quasi-isotropic" (taken from that dragon plate website on page 3), but that's complete BS. In testing, if you end up with a stress that's at an angle just between two major strain direction, you end up with cracks. Carbon is very brittle.
The second problems comes from the construction of the parts themselves. A lot of pre-made composite parts under-perform constantly in tests because of an excess of adhesives. Basically, it means, that the builders use too much epoxy between the fabric. And the strength goes down really quickly when you have too much epoxy. So a lot of parts are closer to a carbon reinforced plastic parts than an actual CF composite construction and will be outperformed by a basic 6061-T6 part. IMHO, if you want a CF structure that will perform, you need some pre-impregnated fabric laid down by pros and then cured in an autoclave. I have never seen good results without that.
What I think is the real biggest problem of composite construction is quality control. For those interested, there is a really good explanation of the problems of composite construction that the Boeing team faced during the development of the X-32 in
this documentary. Even with experts and pros, they had to scrap the first wing they did because of voids founds in the quality inspection. And I think this is the reason why composite construction hasn't really taken off yet. Every part has to be inspected individually. That's expensive, time consuming, requires expensive tools and you have to restart from the beginning if a problem is found. When I was doing testing, we saw a lot of problems (voids, excess epoxy and misaligned fabric) and under-performing parts.
That problem is also present in honeycomb panels. You have panels that either have voids where the skin isn't glued entirely to the honeycomb, or a panel that is overly heavy because they drenched it adhesive.
That's why I think that for an amateurish build, you should go either with a good old steel tubular spaceframe or with an Elise-esque chassis:
While I fully agree about the issues with this technique, given the number of FSAE teams that use the technique, and the fact that it's getting reported in Racecar Engineering very much suggests that those concerns can be dealt with- at least on an autocross scale.
So given your background, any thoughts on how the teams are dealing with the issues you bring up?
In reply to alfadriver :
My post made it sound like CF monocoque is bad, but you can achieve good results. I think that the problem is most people think it will be just like that FG boat they build a few years back. Building a good performing structure is a lot more complex.
1 - You need a lot of experience laying fabric before you get good. I know most FSAE teams completely mess up their first efforts. My old work place used to hire guys with experience building boats, but they all had that bad habit of applying too much epoxy because the part looks good after. But a composite piece with too much epoxy is weak. They wouldn't care about fiber alignment because they used to work with FG mat. And they had a tendency to cut the fabric to large to cut some back later, but they would get lazy. That could screw-up with the load pathways thru the part. I know I wouldn't feel confident trying to make a monocoque that my life would depend on.
2 - To help with the lay-up, you need to CNC cut your fabric. Because fabric tends to move (stretch, shift, etc.) when you cut it by hand, you often get misalignment or size issues.
3 - Tubes, beams, plates and honeycombs were getting better everyday and I am sure they are much better than they were. Since they are mass-produced, you take out the human element and improve quality. I would still like to make my own tests before using a product, but it was getting better. If you can think of a way of making a chassis using these elements, you could have something. We see a lot of trial in the bicycle and FSAE world. The issue is in joining the elements. I haven't seen anything that I really like yet.
4 - Engineers now know that you have to test the properties of the material you will be using before doing any calculations. Most fabrics "age" with time. Most FSAE teams build their monocoque using aged out material that is donated to them by aircraft manufacturers. The amount of epoxy (if you don't go pre-impregnated) will also greatly affect the strength. So building a test panel and verifying the results is a must.
5 - Finding an cheap source of autoclave time is a real hurdle. Most FSAE teams also have their autoclave time given by their sponsor aircraft manufacturer. Which I doubt they would for a small company or an individual building a race car.
6 - Since there isn't a way of fixing a bad part, you don't really care where the issue is, you just have to know if there is one. So testing the finished product is pretty easy. Either some strain gauges or simple torsional rigidity test of your finished chassis will quickly tell you how good it is.
7 - Even if you did everything right, you still have to accept a pretty big variation in the finished product. FSAE teams would get 20-25% variations in their results. You have to be ready to live with that.
It's been 10 years since I was involved in that world, so things might have changed. I am now a casual observer, so YMMV.
In reply to fanfoy :
Specifically to the cutting and folding some kind of plate- either foam or honeycomb sandwiched between layers of CF...
The corners are the huge question, to me. In terms of curing and bonding pre-preg to a cured sheet, are there documents that cover that issue? Then, as you point out, the fiber direction of those patches are quite important...
But it does work.
And given the hugely less amount of specialized tools to do that technique is very interesting- most can find a large sheet CNC machine somewhere near them, whereas a large autoclave is not common.
In terms of testing- one could make some basic shapes and find destructive testing companies near them.... hmmm...
In reply to alfadriver :
The thing with carbon is, once you impregnate it with epoxy... you’re not going to bend it around a corner. It’ll flex, maybe 10deg, and then either fail “plastically” or it’ll be too strong to bend anymore, and then they’ll return to their ‘original flexion state’
That’s probably true for F/G, Kevlar, etc.
So, you’d have to route out both sides of the panel, and some portion of the core to do what you’re talking about. And there’d be a very high probability of the core failing while you’re doing this whole operation.
You’ve got to lay-up carbon in it’s final shape, and knowing what the load path is. You can tie an overhand knot in a “tow” of unimpregnated unidirectional carbon, and at the end you’ll end up with two pieces... it’s that fragile in anything outside of “tug of war” loadings.
In reply to sleepyhead :
I've only read the RE article, not really researching the technique- how are the FSAE teams doing it? Are they using a different material?
I guess I'm mostly interested in replicating what they are doing....
I.......... am Not a Smart Man..........
I have read and understand each Opinion and Fact I have read here so far, I Like where it's Going and Have learned more in Day's than the last 10 years, At this Point Carbon Fiber Scares me and maybe someone else to Death. I want to know Lots More, but do not see it being an Option for the Grassroots Builder, By Myself I don't have the means, But want to see the Process, We have some Trade School Members here that have even Won the Challenge. Any of Ya'll up on North Ave. Interisted, That's the only place close to me that may have the Tools .
The Shattering of Composite Is what scares me the most ,I have heard of Shards becoming Shrapnel,
Saw A Photo once of an all Composite 3/4 scale P-51 I loved it but the thought of it now is Scary, Used a Big Block Chevy BTW.
GTXVette said:I.......... am Not a Smart Man..........
At this Point Carbon Fiber Scares me and maybe someone else to Death.
it's beginning to sound like a black art. I suggest we forget it and go straight to something simple like 3D printing. 
I've no Doubt , With the right tools and materials I can do this, so no Fear.
On the Other hand, I don't have an AutoClave or a CNC in the Shed, some do,
I know it's super lite and super Stiff, If it didn't splinter on impact I would be more Interested.
I do Like Monocoque and tubing like small aircraft and some big ones too, Ilght and strong, and as for Loosening of Fasteners go , That's what Maintanance is about, Also Repairable.
I know Aero is Important But I keep seeing early 60's cars in my head. I don't want this Discussion to end Because I want to Learn,
Let's on Paper Disassimble an older design race car that is Monocoque/tubing , then redo in CF and one in Newer Tech with tubing and mono.maybe honeycomb.
maybe list costs and time to build each Component .I'm sure FSAE have this documented for both design's
In reply to GTXVette :
Here is a dis-assembled aluminum monocoque for you to ruminate over. McRae F5000. Note that it had serious damage, not all revealed yet in this photo.
In reply to alfadriver :
The way to do foam core is usually to have it machine to the desired shape and then to cover it with fabric. I have never personally seen someone using a pre-made flat core and "bending" it. I think that is frown upon even with aluminum honeycomb.
Ok so late to the party and I don't have a working copy of cosmos right now to test this but stop bending this stuff and do slot in hole construction with flat sheet. You could crank out a decent chassis in a day with just a CNC router and nothing else. Would keep your material costs down as well.
You'll need to log in to post.
